Vous êtes sur la page 1sur 313


How Not to Network a Nation

9800.indb 1 6/2/16 3:05 PM

Information Policy
edited by Sandra Braman

The Information Policy Series publishes research on and analysis of signifi-

cant problems in the field of information policy, including state-law-soci-
ety interactions as well as decisions and practices that enable or constrain
information, communication, and culture irrespective of the legal siloes in
which they have traditionally been located. Defining information policy as
all laws, regulations, and decision-making principles that affect any form
of information creation, processing, flows, and use, the series looks at the
formal decisions, decision-making processes, and entities of government;
the formal and informal decisions, decision-making processes, and entities
of private- and public-sector agents that are capable of affecting the nature
of society; and the cultural habits and predispositions that support and sus-
tain government and governance. The parametric functions of information
policy at the boundaries of social, informational, and technological systems
are of global importance because they provide the context for all communi-
cations, interactions, and social processes.

Virtual Economies: Design and Analysis, Vili Lehdonvirta and Edward Cas-
Traversing Digital Babel: Information, e-Government, and Exchange, Alon Peled
Chasing the Tape: Information Law and Policy in Capital Markets, Onnig H.
Policy for Computing Infrastructure: Governance of the Cloud, edited by Chris-
topher S. Yoo and Jean-François Blanchette
Privacy on the Ground: Driving Corporate Behavior in the United States and Eu-
rope, Kenneth A. Bamberger and Deirdre K. Mulligan
How Not to Network a Nation: The Uneasy History of the Soviet Internet, Benja-
min Peters

9800.indb 2 6/2/16 3:05 PM

How Not to Network a Nation

The Uneasy History of the Soviet Internet

Benjamin Peters

The MIT Press

Cambridge, Massachusetts
London, England

9800.indb 3 6/2/16 3:05 PM

© 2016 Massachusetts Institute of Technology

All rights reserved. No part of this book may be reproduced in any form by any
electronic or mechanical means (including photocopying, recording, or information
storage and retrieval) without permission in writing from the publisher.

The documents shown in Figures 4.1, 4.3–4.13, and 4.15–4.19 are held in the person-
al archive of Vera Viktorevna Glushkova and are reproduced here with permission.

This book was set in Stone by the MIT Press. Printed and bound in the United States
of America.

Library of Congress Cataloging-in-Publication Data

Names: Peters, Benjamin, 1980- author.
Title: How not to network a nation : the uneasy history of the Soviet
internet / Benjamin Peters.
Description: Cambridge, MA : MIT Press, [2015] | Series: The information
policy | Includes bibliographical references and index.
Identifiers: LCCN 2015038371 | ISBN 9780262034180 (hardcover : alk. paper)
Subjects: LCSH: Computer networks—Soviet Union—History. | Internetworking
(Telecommunication)—Research—Soviet Union—History.
Classification: LCC TK5102.3.S68 P48 2015 | DDC 384.30947/09045—dc23
LC record available at http://lccn.loc.gov/2015038371

10 9 8 7 6 5 4 3 2 1

9800.indb 4 6/2/16 3:05 PM

Joli Jensen, Michael Schudson, Fred Turner, Gary Browning
Four mentors at four schools

9800.indb 5 6/2/16 3:05 PM

9800.indb 6 6/2/16 3:05 PM

Series Editor’s Introduction ix

Prologue xi

Introduction 1
1 A Global History of Cybernetics 15
2 Economic Cybernetics and Its Limits
3 From Network to Patchwork: Three Pioneering Network Projects That
Didn’t, 1959 to 1962 81
4 Staging the OGAS, 1962 to 1969 107
5 The Undoing of the OGAS, 1970 to 1989 159
Conclusion 191

Acknowledgments 207
Appendix A: Basic Structure of the Soviet Government 213
Appendix B: Annotated List of Slavic Names 215
Appendix C: Network and Other Project Acronyms 219
Notes 221
Bibliography 259
Index 287

9800.indb 7 6/2/16 3:05 PM

9800.indb 8 6/2/16 3:05 PM
Series Editor’s Introduction

Sandra Braman


In this first-ever book-length treatment of early Soviet intelligent network

design history, Benjamin Peters uses uncanniness as a method. He makes
use, that is, of the disorientation that results when the familiar is encoun-
tered in an unfamiliar context, broadening and deepening what we believe Series
that we know about the familiar. This can be a dangerous endeavor. Avatar
designers and others fear the “uncanny valley”—where the nonhuman is so
close to the human that the difference cannot be discerned—because that
is literally too close for viewer/user comfort. “That’s different,” they say in
those cultures with traditional concerns about trolls, those who look like
people but in fact are not.
Historically, the uncanny “other” was supernatural and not necessarily
to be trusted with matters of this world. For Peters, the “other” is the Soviet
Union. What he found is based on original multilingual archival research
and oral interviews with those who were involved in the design processes.
Peters describes what he sees as the capitalist features of the Soviet world
that undermined its networking efforts and what he views as the social-
ist characteristics of the United States that produced the Internet. On the
face of it, this suggests deep contradictions within capitalist and socialist
systems that belie the claimed and apparent differences between the two
blocs. But even those concerned with cybersecurity acknowledge that it can
be difficult to identify the “other” in the network environment. What was
the uncanny valley in this analytical zone?
For those who think about the information economy, differences between
the East and West are indiscernible. Cristiano Antonelli’s (1992) seminal
insights into the nature of the information economy, in which coopera-
tion and coordination are as important as—or more important than—com-
petition for long-term economic success were inductively developed from

9800.indb 9 6/2/16 3:05 PM

x  Series Editor’s Introduction

detailed studies of the practices and activities of transnational corporations

on both sides of the iron curtain (many funded by the unfortunately short-
lived United Nations Center on Transnational Corporations). What Peters
presents as counterintuitive actually provides further evidence of the transi-
tion to a global information economy in which ideological differences may
still provide motivations but not explanations. Work of this kind, which
looks across political environments, is particularly valuable as we struggle
to make policy for a world in which network politics is genuinely global
even though state-centric geopolitical distinctions remain.
Theoretical pluralism has been familiar since the 1980s, but on reading
Peters one suddenly realizes that most of those who take such an approach
tend to prefer particular types of causal probability even as they roam across
theories and disciplines. Peters is not only interdisciplinary but also travels
across the levels and qualities of the likelihood that any given causal factor
will be determinative in a given circumstance. In this history of early Soviet
network design efforts, Peters ranges from unpacking institutional rigidi-
ties that did successfully shape knowledge production and use to focusing
attention on contingencies that can radically affect ultimate outcomes. His
heterarchical approach to policy analysis importantly reminds us of the
need to examine the interplay among decision-making processes as well as
among players. And Peters returns again and again to the centrality of ideas
in policymaking, devoting a full chapter to the history of cybernetics in the
Soviet Union during the period covered.
Oddly, according to the OED, the notion of the “uncanny” came into
written use a century before the word “canny” was seen. This may be an
artifact of the processes by which materials survive, but it is still interesting.
Peters’s multilingual archival research and oral interviews with individuals
involved in the Soviet efforts have yielded a picture of network conceptu-
alization and decision-making processes fascinating not only in their own
right but also for what they offer to those who study and live with intel-
ligent networks in other parts of the world. We are the other, in the global
network. With this book, Peters deepens our ken of networks—a funda-
ment of information policy since at least the 1830s and the telegraph—and
brings their study into the next generation.

9800.indb 10 6/2/16 3:05 PM


The seeds of this book were first planted as I stood on the left bank of
the Volga River in Balakovo, Russia, one evening in the spring of 2001.
Balakovo, where I was living for several months doing volunteer service,
was a pleasant city of roughly 200,000 people who were struggling through P
the economic depression that was sweeping Russia’s rust belt. As I reflected
on my picturesque surroundings—green trees, rolling hills, and the setting
sun’s reflections on the river—I sensed that something was out of place.
The peculiar features that were visible on the horizon, backlit by the sunset,
belonged to the Saratov hydroelectric dam, one of the world’s hundred larg-
est dams by output and stretching over 1,200 meters in length to form the
enormous Saratov reservoir. The city of Balakovo also is home to a thermal
heat power plant and a nuclear power plant with four working nuclear reac-
tors (the construction of two other nuclear plants was suspended in 1992).
If local rumors were to be believed, Balakovo once boasted secret Soviet
military factories, one of which produced a material for the cosmonautic
industries that was so tough that napalm balled up and rolled off it. This
peculiar pairing of natural scenery and outsized industrial infrastructure
struck me on the riverbank that evening. What force of imagination and
statecraft, I puzzled, could have decided to graft such hulking industry onto
such a remote city—and why would it do so? Thus began my interest in the
outsized infrastructural imagination of Soviet planners.
Six years later, in 2007, those seeds sprouted into the question driving
this book. As a doctoral student at Columbia University, I wanted to learn
more about the international sources of the information age, a topic that
first crystallized for me in Fred Turner’s graduate seminar on “Computers,
Information Ideology and American Culture since World War II” at Stan-
ford University in 2005. If, to gloss Whitehead, all philosophy begins as a

9800.indb 11 6/2/16 3:05 PM

xii Prologue

series of footnotes to Plato, then this book began with an obscure footnote
in Flo Conway and Jim Siegelman’s popular biography of Norbert Wiener.
As I was rereading the book’s references one evening in 2007, I stumbled on
a passing reference to a declassified, Freedom of Information Act–recovered
1962 Central Intelligence Agency report about a new Soviet initiative to
develop a native “unified information network.”1
That footnote triggered a question that was so tenacious that I had to
write this book to shake it: why were there no Soviet developments com-
parable to the ARPANET in the 1960s? It made sense that, at the height
of the cold war technology race, Soviet cyberneticists would try to build a
“unified information network”—and yet I knew nothing about their efforts
or outcomes. I was hooked. What had happened? Why was there no Soviet
Over the next eight years, the question drew me to archives and inter-
views in Moscow and Kiev. After spending a year exhausting the available
leads, literature, and FOIA requests available from New York, I traveled to
begin archival work in Moscow, although initially this proved a dead end.
Marshall McLuhan once quipped that the first thing a visitor needs to know
about Russia is that there are no phonebooks.2 His point is that a foreigner
in Russia needs to have contacts already in place. (Or as the Finns say: in
Finland, everything works and nothing can be arranged. In Russia, nothing
works but everything can be arranged.) And so, with all the tools but none
of the social network, I found myself shuffling through dusty documents
that were lit by a single flickering light bulb in Moscow archives. Then in
2008, good fortune smiled when, while chasing down references to Niko-
lai Fedorenko and Viktor Glushkov in Moscow, I began a correspondence
with the Massachusetts Institute of Technology historian Slava Gerovitch,
who emailed me from Cambridge a draft of his article “InterNyet: Why the
Soviet Union Did Not Build a Nationwide Computer Network” that became
the basis for this book.3 Gerovitch also put me in touch with key contacts
in Kiev, and my rapidly expanding social network led to dozens of inter-
views and contacts, out-of-the-way archives (including stacks of papers in
the closet of an abandoned office), and unprecedented access to historical
materials over years of research and writing. On the surface, this book is
about why certain computer networks did not work in the Soviet Union,
but the story turns on the basic fact that social networks in the region have
long operated according to their own rhythms and reasons.
Writing this book has proven to be a valuable learning process. When
I set out in 2007 to study early Soviet networks, I had a vague sense that
the resulting scholarly work would intersect media and communication

9800.indb 12 6/2/16 3:05 PM

Prologue xiii

studies, the history of science and technology, and social thought that
informed information policy discussions, but I did not anticipate the work
in institutional and historical economics, the sociology of economics, and
organizational theory that the story required. Least of all did I imagine that
this story would throw me headlong into a study of Soviet bureaucracies. It
is my hope that this work will lighten some of that burden for the patient
reader. In the end, this book should be understood primarily as an inter-
disciplinary work of synthesis driven by a fascination with the relationship
between communication technology and people. I have tried to write for
the media and technology scholar as well as the general-interest reader,
although the book draws on history, area studies, and social commentary
to inform the emergent subfield of network studies in information policy
as well. Like all these fields, its primary orientation is not to a single dis-
cipline but to the scholarly enterprise of making strange modern network
culture, a technique that the Soviet critic Viktor Shklovsky first popularized
as ostranenie, or “defamiliarization.”4 It seeks to offer what historian Peter
Brown calls “salutary vertigo” or a disorientation that clarifies the foreign-
ness of a modern networked culture that was once thought familiar.5 To
do so, this work seeks to separate readers from hidden assumptions about
modern networks and the social, technological, political, and economic
conditions that organize and are subsequently organized by it. For me, this
book began as an essay on the forgotten origins of computer networks in
the Soviet Union and ended up being about much more, including a cau-
tionary tale in the annals of technological innovation and a critical reflec-
tion on the assumptions steeping the current network world.

9800.indb 13 6/2/16 3:05 PM

9800.indb 14 6/2/16 3:05 PM

I n

There is much which we must leave, whether we like it or not, to the un-“scientific”
narrative method of the professional historian.
—Norbert Wiener, Cybernetics, or Control and Communication in the Animal and the
Machine, 1948, concluding line I n

The Soviet Union was home to hidden networks. The story told here about
those networks hangs on a hook that is unfamiliar to most readers and
scholars—the Soviet Internet. At first glance, pairing the Internet and the
Soviet Union appears paradoxical. The Internet first developed in America
and became popular only after the Soviet Union collapsed. The Internet
suggests to general readers open networks, flat structures, and collaborative
cultures, and the Soviet Union signals censored networks, hierarchies, and
command and control cultures. What, then, could the phrase Soviet Internet
possibly mean?
The central premise of this book holds that there was once something
that we might think of as the Soviet Internet. Between the late 1950s and
the late 1980s, a small group of leading Soviet scientists and administra-
tors tried to develop a nationwide computer network that was designed for
citizen communication and sweeping social benefits. This book is about
their story. At the height of the cold war technology race, the Soviet Union
was awash in intelligence about contemporary Western initiatives, includ-
ing the Semi-Automatic Ground Environment (SAGE) project at the U.S.
Department of Defense. The Soviet state had all the necessary motives,
mathematics, and means to develop nationwide computer networks for the
benefit of its people and society. This book also ventures analysis on why,
despite pioneering national network projects from the most promising of
scientists and administrators, the Soviet state proved unable and unwilling
to network its nation.

9800.indb 1 6/2/16 3:05 PM

2 Introduction

This much is clear: the Soviet Union never had the Internet as it is
known today.1 Rather, in the early 1960s, Soviet cyberneticists designed
the most prominent of the network projects examined here—the All-State
Automated System (OGAS)—with the mission of saving the entire com-
mand economy by a computer network. Their elaborate technocratic ambi-
tion was to network, store, transmit, optimize, and manage the information
flows that constituted the command economy, under the guidance of the
Politburo and in collaboration with everyday enterprise workers, managers,
and planners nationwide.
The historic failure of that network was neither natural nor inevitable.
Its story is one of the lifework and struggles of often genius cybernetic sci-
entists and administrators and the institutional settings that were tasked
with this enormous project. The question deserves a sympathetic and rigor-
ous examination of the Soviet side of the story. Why did Soviet networks
like the OGAS not take root? What obstacles did network entrepreneurs
face? Given unprecedented Soviet investments and successes in mathemat-
ics, science, and some technology (such as nuclear power and rocketry),
why did the Soviet Union not successfully develop computer networks that
were capable of benefiting a range of civilian, economic, political, social,
and other human wants and needs? How might we begin to rethink our
current network world in light of the Soviet experience?
I propose that the primary reason that the Soviets struggled to network
their nation rests on the institutional conditions supporting the scientific
knowledge base and the command economy. Those conditions, once exam-
ined, challenge conventional assumptions about the institutions that build
open, flat, and collaborative networks and thereby help recolor the cold
war origins of the information society. It is a mistake, as the standard inter-
pretation among technologists and some scholars have it, to project cold
war biases onto this history. Our networked present is the result of neither
free-market triumphs nor socialist state failures.
That said, let us begin with a slight twist on the conventional cold war
showdown: the central proposition that this book develops and then com-
plicates is that although the American ARPANET initially took shape thanks
to well-managed state subsidies and collaborative research environments,
the comparable Soviet network projects stumbled due to widespread unreg-
ulated competition among self-interested institutions, bureaucrats, and
other key actors. The first global civilian computer networks developed
among cooperative capitalists, not among competitive socialists. The capi-
talists behaved like socialists while the socialists behaved like capitalists.
In the process of examining and elaborating on that plain statement
about the cold war history of networks, this book describes two intersecting

9800.indb 2 6/2/16 3:05 PM

Introduction 3

approaches to larger questions of social control and change—one institu-

tional and the other technological. The first approach looks at the context
of Soviet institutions and political bodies that were preoccupied with both
the paperwork and the power brokerage behind the socialist command econ-
omy. The question of how to organize economies, especially but not only the
Soviet command economy, is shown to be political before it is economic. The
second approach accounts for the attempts of Soviet cyberneticists to build a
computer system over a period of about thirty years from 1959 to 1989 that
would control in real time the economy’s problems. The two approaches—
political economy and computing technology—combine and play out here
on the common stage of Soviet cybernetics, a midcentury discipline that was
interested in systematizing all organization problems with computing tech-
nology. The result is a tragic story that addresses questions that are central
to the history of technology and global media theory: what makes the same
technology take shape differently in different contexts?
To explore that tragedy, the book sets up the dramatic potential of a
networked command economy, the loss of that potential in the hands of
the state, and a critical reclamation for contemplation, reflection, and con-
temporary instruction. The limitations of this work’s scope are also clear.
Although it focuses primarily on the cybernetics and economic concerns
besetting Viktor Glushkov and his Kiev-based OGAS team between 1959
and 1989, the setting is broader, including the military, industrial, and aca-
demic complexes that stretched from the seat of power Moscow to other
cities, including St. Petersburg to the north and Akademgorodok (a science
city that was nestled deep in Siberia) three thousand kilometers to the east.2
The book also seeks to comment on the Soviet Union as a perceived state
of exception on the global geopolitical stage. As one pole in the global
cold war, the Soviet state stood unrivaled among socialist states in terms of
international military and political influence.3 In their search for a balance
of focus and breadth, historians of science and technology have called for
midpicture history, or a case study drilled deeply to explore intersecting
historical subdisciplines (not entirely unlike Robert Merton’s middle-range
theory). This book is not a midpicture history, although I hope its best
moments may model how media history and theory can move in tandem
with information science and technology. In its most ambitious moments,
this book offers a synthesizing commentary (in the premodern sense of the
central genre of scholarship, not derivative status) about the sources of the
modern network age.4
This book seeks to complicate the popular memory of the Soviet Union—
its heady promises of socioeconomic justice as well as its parade of horribles,

9800.indb 3 6/2/16 3:05 PM

4 Introduction

including authoritarian abuses, violence, and a cumbersome state hierarchy

that subjected its citizens to political oppression and information censor-
ship. It examines the Soviet command economy, which proved inflexible
to the fluctuating demands of the emerging global network economy and
eventually imploded on itself. Some readers may feel that the Internet and
the Soviet Union seem to be fundamentally opposed information projects:
one is a salvific vehicle for the invisible hand of modern-day commerce,
and the other is remembered for its dead hand; one led to the knowledge
explosion that is Wikipedia, and the other, to the nuclear catastrophe at
Chernobyl; one produced Linux, and the other, the Lada; one is a haven for
technoenthusiasts, libertarians, and free-speech absolutists, and the other,
the whipping boy for the same. But I seek to bring to English-language
readers the story of the Soviet computer network in its own terms. Given
that the story is not singular, my emphasis is on relating the untold story
of the All-State Automated System project and its research network led by
the mathematician Viktor Glushkov in Kiev (the current capital of Ukraine)
between 1959 and 1989. The case study arrives couched in commentary
that seeks to upend and move beyond residual binary narratives about the
cold war origins of the current networked age.5
The internal historical setting for the tragic tale begins with the tur-
bulent grab for power that followed the death of Joseph Stalin in 1953
and stretches through the halting internal unraveling of that power in the
1980s. There was an unusual contender for filling the political vacuum left
by Stalin’s passing. To the scientists under study here, Stalin’s best replace-
ment was no person at all but rather a technocratic conviction that com-
puter-aided governance could avoid the past abuses of its strongman state.
The All-State Automated System was a utopian vision of a distinctly state
socialist information society as well as, closer to home, a familiar story of
how bright men and women struggle to employ both might and machines
in the service of social justice and greater public goods.6
A thin line sometimes separates tragedy from comedy. Backlit by reflec-
tions on cold war political economic orders, the fickle muses of historical
contingency staff this drama. For example, family preferences for warmer
weather ended up shifting the centers of scientific development, empty
chairs at crucial meetings sank decade-long campaigns, informal whims of
power shipwrecked careers and perhaps countries, basic notational systems
(not sophisticated algorithms) revolutionized long-term strategic thinking
(and Soviet chess), and countless other details rained down via informal
bureaucratic actions on the Soviet knowledge base. All these and others
blur the comic and tragic elements. The Soviets could have developed a

9800.indb 4 6/2/16 3:05 PM

Introduction 5

network contemporary to the ARPANET, and yet they did not. What makes
this story tragic is not that the Soviet political, economic, and technological
networks collapsed but that the deeper problems that beset the USSR have
been transformed but have not disappeared. The twenty-five years follow-
ing the collapse of the Soviet Union have reaffirmed that Russia, although
no longer in a superpower showdown with the West, remains anything but
a negligible actor on the global stage and that the patterns of its state gov-
ernance are much older than the post-Soviet transition. By triangulating
across the central Soviet-American cold war axis to emphasize Ukrainian
and other liminal people and places, this book aims to help readers rethink
residual cold war misunderstandings in popular network and digital media
discourse while simultaneously showcasing the institutional tensions at the
heart of modern-day networked practices, policies, and polities.

The curtain parts on two anecdotes about Soviet networks. The first intro-
duces the central story, and the second marks the limit of that story. In late
September 1970, a year after the ARPANET went online, the Soviet cyber-
neticist Viktor Glushkov boarded a train from Kiev to Moscow to attend
what proved to be a fateful meeting for the future of what we might call
the Soviet Internet. On the windy morning of October 1, 1970, he met with
members of the Politburo, the governing body of the Soviet state, around
the long rectangular table on a red carpet in Stalin’s former office in the
Kremlin. The Politburo convened that day to hear Glushkov’s proposal and
decide whether to build a massive nationwide computer network for citi-
zen use—or what Glushkov called the All-State Automated System (OGAS,
obshche-gosudarstvennyi avtomatizirovannaya system), the most ambi-
tious computer network project of its kind in the world at the time. OGAS
was to connect tens of thousands of computer centers and to manage and
optimize in real time the communications between hundreds of thousands
of workers, factory managers, and regional and national administrators.
The purpose of the OGAS Project was simple to state and grandiose to imag-
ine: Glushkov sought to network and automatically manage the nation’s
struggling command economy.
What transpired in Stalin’s former office that day enters into the story.
Throughout this (and perhaps all) history, the messy details often matter
most. In this case, two crucial chairs in that committee room were empty
on that particular day due to the contingencies of the calendar and compet-
itive bids for power. This book’s analysis will note how pesky details often
reveal hidden patterns of institutional (mis)behavior that structure and
reshape the interests of public actors, organizations, and even economic

9800.indb 5 6/2/16 3:05 PM

6 Introduction

and social relations. Taken together, the history and analysis of the OGAS
and related attempts to network and command the Soviet economy tell a
story with consequences for the history of cold war computer networks and
our understanding of the current networked world that emerged from the
cold war itself.
The second anecdote took place not far from Glushkov’s fateful meeting
in the Kremlin. Here, in a top-secret chamber in a cement bunker, or shariki
(“spheres” or “globes”), buried deep underground somewhere outside of
Moscow, was a very different kind of computer network. In that small room,
a few uniformed personnel sat before flickering computer screens that were
powered by an independent generator purring audibly nearby but out of
sight. The single closed door was of reinforced metal with a self-locking
mechanism, and behind it a long ladder ascended into a network of under-
ground tunnels overhead. The chairs were bolted to the floor and pivoted
to allow the military officers to review a control panel lined with informa-
tion displays—satellite data and security camera feeds, telephone and radio
signals, Geiger counters and seismographs, and other instruments for mea-
suring the world above. These men sat at their consoles, operating as cogs
in a larger sociotechnical machine. They were trained so that if or when
the time arrived, they would observe the sensors, orient and input certain
coordinates and a timetable, flip switches, and press a button that would
end the world in a nuclear Armageddon.
This is Dead Hand, the semiautomatic nuclear-defense perimeter sys-
tem that was first installed in the late Soviet Union. The details above are
mostly pure invention, and yet the network system is real. Formally called
Systema “Perimetr,” the perimeter system was imagined under Brezhnev as
a fail-deadly deterrence mechanism for ensuring second-strike capacity in
the nuclear cold war.7 These men—not unlike the U.S. workers who staffed
the Emergency Rocket Communication System from 1961 to 1991—sat in
the top-secret underground command-and-control center of their nation’s
perimeter system. The data were fed into computer consuls to confirm
whether the enemy had struck first. If an American military strike effec-
tively disabled the regular Soviet command-and-control military leader-
ship above ground from swiftly retaliating, then the strategy maintained
that the Dead Hand would stand ready to trigger “a spasm of destruction.”8
After the national computer network system was activated, it would put on
alert nuclear-tipped intercontinental missiles that were stored thousands
of miles away. The red button, once pressed, would launch a massive retal-
iatory nuclear strike, enacting swift revenge at a global cataclysmic scale.
Behold the apocalypse—delivered by national network.

9800.indb 6 6/2/16 3:05 PM

Introduction 7

This book is about civilian networks, not military networks. This is a

deliberate choice. I choose to emphasize public networks because a network
built for every Soviet worker still speaks to the popular and scholarly imagi-
nation of our current socially networked world in ways that closed military
networks do not, although, as we will see, the military’s relationship to
technological innovation backlights the whole stage of cold war science.
A sideways look at some of the discourse about online commerce today
proposes the enduring relevance of the Soviet socialist revolution that was
consummated a century ago. Both the Internet and the Soviet command
economy promise the revolutionary realization of the means for socialist or
collectivist production on a mass scale. In the rhetoric of networking col-
lective consciousness and crowd-sourced collaboration, we see the unlikely
alliance of Wired editor Kevin Kelly’s hive mind, open-source software pro-
moter Eric Raymond’s bazaar, and Marxist revolutionary Leon Trotsky’s
collective farm.9 Long before Internet enthusiasts were around, Soviet
enthusiasts were promising that workers (users) could meet the needs of
the masses (crowds) through collective modes of resource sharing and col-
laboration (peer-to-peer production).
Few, if any, contemporary scholars recognize these concerns as funda-
mental to our modern network culture, and yet they persist in coloring
views of both past and future. This is no accident: the concurrent emer-
gence of cyberspace and post-Soviet affairs entered scholarly and popular
discourse at the tail end of the previous century. For example, sociologist
Manuel Castells has developed an extensive argument detailing how the
Soviet Union failed to enter the information age, which this book is in
some ways a sideways response to, and legal scholar Lawrence Lessig used
his experience observing the rapid deregulation and privatization in post-
Soviet economic transition in the early 1990s as a formative analog for what
he felt was an equally disastrous attitude about the supposed unregulability
of cyberspace common in the late 1990s.10 Since then, scholars have recog-
nized that the summary experiences of perhaps the last two great informa-
tion frontiers of the twentieth-century—the rise of post-Soviet economic
transition and the Internet—present not, as Francis Fukuyama infamously
claimed, the end of history so much as a new chapter in it. Leading cyber
legal scholar Yochai Benkler has argued for a middle way by observing how
online modes of “commons-based peer production” sustain capitalist profit
margins through collectivist forms of reputational altruistic communities
that do not depend on individual self-interest.11 From the final chapters
of Soviet history, we may begin to observe and puzzle through the peren-
nial fact that, for many Western technologists and scholars, the promise of

9800.indb 7 6/2/16 3:05 PM

8 Introduction

socialist collaboration shines brightest online today—a promise that the

Soviet OGAS designers were among the first to foresee.
None of the conditions—technological, sociological, economic, or oth-
erwise—for the flourishing of computer networks are necessarily as we
may think. As Melvin Kranzberg’s first law of technology holds, technol-
ogy is neither positive, negative, nor neutral.12 The same holds for society
and economy. By looking at failed network projects, I seek to flip science
anthropologist and philosopher Bruno Latour’s aphorism that technology
is society made durable. We observe in the collapse of the Soviet network
projects a lesson for humans who live in a fragile world: society too is tech-
nology made temporary.13 The Soviet experience with networks reminds us
that although computer networks are prospering today, our modern social
assumptions about those networks are no more inevitable or permanent
than those of the Soviets. Our current beliefs about networks will pass. This
book looks to take in a new direction what science and technology scholars
Geoffrey C. Bowker and Leigh Starr have called an “infrastructural inver-
sion”: looking closely at the alternative setting of a Soviet networked soci-
ety can shake up a modern mental infrastructure that makes the current
networked environment appear natural and necessary.14 Sometimes the
best way to see something is to look away from it. The French revolution,
as historian Eric Hobsbawm has noted, did not become the French revolu-
tion until it was seen in the context of the British industrial revolution
and the revolutions of 1848.15 We stand to apprehend the current network
transformations better by placing the past in the context of a wider world.
By exploring the pathway that was once taken and then abandoned in cold
war networks, I hope to help unsettle, broaden, and deepen our imagina-
tion for the possibilities that gave rise to the modern networked media
The literature on which this book builds is growing. Above all, this
book builds on the historical foundation that was laid by the pioneering
works of historians of Soviet science, Slava Gerovitch and Loren Graham.16
Slava Gerovitch’s article “InterNyet: Why the Soviet Union Did Not Build
a Nationwide Computer Network,” which he shared with me in draft form
while I was independently pursuing the Soviet Internet story in archives in
Moscow, jumpstarted this history with a treasure chest of scholarly leads.
His work has opened many windows into the Soviet history of science and
its associated social problems. The literature in English on the midcentury
development of computer networks—by leading scholars such as Janet
Abbate, Finn Burton, Paul N. Edwards, Fred Turner, and Thomas Streeter—
also includes works that examine the creative communities, institutional

9800.indb 8 6/2/16 3:05 PM

Introduction 9

innovation and setbacks, cold war tensions, and Western internal politics
that backlight this particular case study.17 This work attempts to help inter-
nationalize the core insights of this sociologically sensitive body of analysis
into the people and places that shape networks.
The literature also teaches that the significance of the global spread of a
social network often precedes, exceeds, and coevolves with that of any spe-
cific technological network. To borrow a line from Elihu Katz, international
communication networks precede national computer networks.18 Along
these lines, historian of technology Eden Medina’s Cybernetic Revolutionar-
ies: Technology and Politics in Allende’s Chile advances a seminal history and
analysis of early technological and political attempts to network another
socialist state during the cold war. Her close and careful analysis of the peo-
ple involved in the creation of Project Cybersyn (especially 1971 to 1973)
reveals how the significance of technological projects carries beyond and
exceeds that of specific network projects.19 Her work, together with other
recent scholarship on international cybernetic movements, helps outline
the central cast of characters in this book.20 This cast was not selected exclu-
sively from cybernetic scientists or administrators. Rather, the characters
are drawn from what I call the “knowledge base” of the Soviet Union—the-
oretical and applied scientists, their laboratories and research centers, stu-
dents in universities, administrators in the academies of science, state office
bureaucrats, generals in the Ministry of Defense, ideologues and censors
in the scholarly and public press, the secret police, functionaries, officials,
midlevel managers, members of the Central Committee of the Communist
Party, and others whose careers depended on the management, manipula-
tion, and representation of knowledge as an intellectual, institutional, and
innovative product.21
Finally, a practical note about language. All translations from Russian
and Ukrainian into English are my own unless otherwise noted. In translat-
ing, as Stephen Jay Gould says, “we reveal ourselves in the metaphors we
choose for depicting the cosmos in miniature.”22 This is true of the transla-
tion process as a way of trying to bring separate languages into resonance.
Sometimes words can be translated straightforwardly. For example, this
work, an interdisciplinary exercise in the emerging field of network studies,
seeks to articulate a fluid discourse around the central term network. The
term network, like other keywords in digital discourse, packs more mean-
ing than is usually seen and has roots in the textile industry of lacework,
like the Jacquard loom behind computer programming techniques (there
may be more silk than silicon to the information age). The Russian term
set’ maps fairly well onto my three English uses of the term network—(1) a

9800.indb 9 6/2/16 3:05 PM

10 Introduction

technical communication network understood as interlinked digital, elec-

tronic, telephonic, or other channels of communication; (2) the complex
sociotechnical assemblage of heterogeneous relations that link people,
institutions, and the administration of markets, states, and other actors in
everyday life; and (3) an abstract organizational mode that maps the link-
ages between any set of objects, such as graph theory in mathematics.23
Although all of these meanings are in play here, what we assume to be a
relatively settled term today behind the concept of network (set’) took up in
Soviet discussions an even wider set of terms such as base, complex, cluster,
and most characteristically for computers connected over distances, system.
At other times, Russian terms reveal their own world in how they resist
easy translation. I occasionally retain, for example, the early Soviet term for
computer, “the automatic high-speed electronic calculating machine” (avto-
maticheskaya byistrodeistvuyushchaya elektronicheskaya schyotnaya mashina
and its various shortenings) for its splendidly descriptive bulk that signals
perhaps the most elegant definition of new media I know: new media are
those media we do not yet know how to talk about.24 The probability theo-
rist Aleksandr Ya. Khinchin revealingly renders what is known in English as
“queuing theory” (used by information theorists to describe how data pack-
ets wait in line) as “mass-service theory” (teoria massovogo obsluzhivaniya) in
Russian.25 Sustaining the anthropological gaze requires depicting the vari-
able sets of cultural, social, and political values in comparative relief with
the network elements that are all too familiar in modern culture, which I
have attempted to do here whenever relevant.
I have also tried to write with the conviction that plain language packs in
its own insights. By proposing for further examination that the first global
civilian networks took shape thanks to capitalists behaving like socialists,
not socialists behaving like capitalists, I understand the terms capitalism
and socialism in the ordinary way. I define capitalism as the order of the
market economy, where economic actors act independent of the state, pri-
vate property rights are reasonably secure and dominate most enterprises,
prices and trade are predominantly free, state subsidies are limited, and
transactions mostly monetized. Socialism, by contrast, is an economic order
of the command economy where the opposite can usually be expected,
although with its instinct to communism operating according to the moral
and political principle “from each according to their abilities, to each
according to their needs.”26 The argument here depends not on collapsing
that definitional divide but on revealing how that ordinary understanding
falls short of describing mixed constellations of competitive and collabora-
tive practices—public-private and state-market formations that belie and

9800.indb 10 6/2/16 3:05 PM

Introduction 11

tweak our sense of these opposing economic orders. Evidence complicates

the tidiness of ideas. This is a conventional a priori to foundational work
in general scholarship and in institutional economics, which look to the
complexities of behavior and scale them toward understanding the unpre-
dictable behaviors of modern state and market relations.27
At other times, new phrases have been introduced to familiarize readers
with a foreign context. I have attempted to cast a critical eye on all source
materials, and the careful act of weighting and arranging evidence has
pressed on my work its own brand of insight and argument. For example,
after observing the extraordinary lengths to which Soviet scientists went to
promote economic reform with networks, I settled on the phrase network
entrepreneur to cast a new light on the dynamics of the knowledge base in
Soviet science and technological innovation. This word choice might seem
misplaced because the Soviet knowledge base appears at first glance to carry
none of the cultural or conceptual weight of venture capital, investment
risk, and inherited responsibility for an enterprise that typically is associ-
ated with the modern English term entrepreneur. And yet the Soviet Internet
makes a fitting case study in the global history of technology entrepre-
neurs, from Thomas Edison to Steve Jobs to Sergei Brin. That history has yet
to be written, although when it is, it will feature an international species of
actors, among them Soviets, who were prone to repeat bold slogans before
proceeding by bolder failures.28 Those who are uncomfortable applying a
capitalistic term to comparable socialist practices may do well to recall that
the English entrepreneur is already on loan from the French.


This book proceeds in five roughly chronological chapters. Chapter 1 intro-

duces the global consolidation and spread of cybernetics as a midcentury
science in search of self-governing systems from World War II to the mid-
1960s. It also notes that cybernetics articulated internationally distinct
scientific dialects to try to harness a range of different information sys-
tems—including biological, mechanical, and social—under one umbrella
science. The term heterarchy is introduced as a cybernetic term for com-
plex networks with multiple conflicting regimes of evaluation in opera-
tion at the same time. Also looked at are the mind and its neural networks
(including the brain and the nervous system) as an international analogy
of choice for thinking about national networks. Then the chapter examines
the historical backdrop of the sequential rejection, adoption, adaptation,
and mainstreaming of cybernetics in the 1950s and 1960s Soviet Union,

9800.indb 11 6/2/16 3:05 PM

12 Introduction

against which the central tragedy of the remaining chapters and cast of
characters unfolds.
Chapter 2 examines the emergence of economic cybernetics in the late
1950s and early 1960s as a field that was closely allied to mathematical
economics and econometrics yet peculiar in its implications in the interna-
tional sphere of Soviet intellectual and political influence. It also outlines
and describes the basics behind the command economy and the tremen-
dous coordination problems that the Soviet state and competing schools
of orthodox, liberal, and cybernetic economists all agreed needed to be
addressed and reformed in the early 1960s. A few sources of the organi-
zational dissonance, including heterarchical networks of institutional
interests, that was underlying the Soviet command economy and its state
administration are also introduced.
Chapter 3 chronicles the first three aborted attempts to network the
Soviet nation. The first was Anatoly Kitov’s pioneering proposal in the fall
of 1959 to build a nationwide computer network for civilians on preexist-
ing military networks. The resulting show trial removed him, the first Soviet
cyberneticist and a star military researcher, from the military. The second
attempt was the short-lived technocratic proposal by Aleksandr Kharkevich
in 1962 to build a unified communication system for standardizing and
consolidating all communication signals in the Soviet Union. And the third
attempt was the simultaneous proposal by N. I. Kovalev for a rational sys-
tem for economic control using a nationwide web of computer networks.
Brief attention is paid to the historical concurrence of cold war networks,
including a caution against the cold war preoccupation to overvalue claims
to being historically “first” in and outside of Soviet science.
Chapter 4 introduces the most ambitious and prominent of Soviet net-
work projects—the All-State Automated System (OGAS)—and its primary
promoter and protagonists, the cyberneticist Viktor M. Glushkov, whose
stories are brought together for the first time in English. This chapter details
what is known about the sweeping theoretical and practical reach of the
OGAS Project between 1962 and 1969, its vision for an economy managed
by network, and the institutional landscape that evolved in support of that
initial project proposal in the 1960s. It also presents snapshots of both the
playful work (counter)culture and informal institutional obstacles that
began to preoccupy two of the most prominent research institutes for eco-
nomic cyberneticists—Nikolai Fedorenko’s Central Economic-Mathematical
Institute (CEMI) and Viktor Glushkov’s Institute of Cybernetics—in the same

9800.indb 12 6/2/16 3:05 PM

Introduction 13

Chapter 5 chronicles the slow undoing of the OGAS between 1970 and
1989. Neither formally approved nor fully rejected, the OGAS Project found
itself (and proposals to use computer-programmed networks to plan social
and economic resources, including those by the chess grandmaster Mikhail
Botvinnik) stalemated in a morass of bureaucratic barriers, mutinous min-
istries, and institutional infighting among a state that imagined itself as
centralized but under civilian administration proved to be anything but. By
the time that Mikhail Gorbachev came to power, Glushkov had died, and
the political feasibility of technocratic economic reform had passed. This
chapter frames how hidden social networks unraveled computer networks.
The conclusion reflects on and complicates the plain statement that is
the conceit of this book—that the first global computer networks began
among cooperative capitalists, not competing socialists. Borrowing from
the language of Hannah Arendt, it recasts the Soviet network experience in
light of other national network projects in the latter half of the twentieth
century, suggesting the ways that the Soviet experience may appear uncom-
fortably close to our modern network situation. A few other summary
observations for scholar and general-interest reader are offered in close.

9800.indb 13 6/2/16 3:05 PM

9800.indb 14 6/2/16 3:05 PM
1  A Global History of Cybernetics


I am thinking about something much more important than bombs. I am thinking

about computers.
—John von Neumann, 1946

Cybernetics nursed early national computer network projects on both sides

of the cold war. Cybernetics was a postwar systems science concerned with
communication and control—and although its significance has been well
documented in the history of science and technology, its implications as a
carrier of early ideas about and language for computational communica-
tion have been largely neglected by communication and media scholars.1
This chapter discusses how cybernetics became global early in the cold war,
coalescing first in postwar America before diffusing to other parts of the
world, especially Soviet Union after Stalin’s death in 1953, as well as how
Soviet cybernetics shaped the scientific regime for governing economics
that eventually led to the nationwide network projects imagined in the late
1950s and early 1960s.
The term cybernetics evades easy definition. Today there are still more
self-identified cyberneticists in the world than available definitions of the
field, although the first tally is dropping as the second tally creeps slowly
upward. In the English-speaking information science research environ-
ment, cybernetics failed to cohere as an institutionalized field, a fact that
partially explains the inability of specialists to agree on a definition for the
field. And yet the definitions are no easier in the territories of the former
Soviet Union, where cybernetics did take root and still enjoys institutional
recognition fifty years later. To this day, the definition puzzle holds: the
postwar science remains a rich subject for critical inquiry precisely because
it has escaped a clear-cut characterization.

9800.indb 15 6/2/16 3:05 PM

16  Chapter 1

Since the mid-1940s, cybernetics’ themes of communication and control

in computational biological, social, and symbolic systems have inspired
and bedeviled researchers across the natural sciences, social sciences, and
humanities. Accounts have identified cybernetics as a science of commu-
nication and control, a universal science, an umbrella discipline, a Mani-
chean science, and a scientific farce founded on sloppy analogies between
computers and human organisms.2 Against its interdisciplinary backdrop of
a computer-compatible formulation of communication, scores of scientists,
philosophers, and policy makers advanced the midcentury computer as a
tool for modeling systems and specifically for the regulation of information
flows and behavior in the animal, the machine, and society. In addition
to computer modeling, it gathered together preexisting concepts such as
feedback loops in control systems, cooperative human-machine relations,
and some foundations for the network design of digital computing. In the
information sciences, it formalized midcentury mind-machine analogies
that continue to animate some corners of contemporary artificial intelli-
gence research. In the hands of polymaths such as Norbert Wiener, Warren
McCulloch, and Donald MacKay, the technical and technocratic insights
into a summary set of cybernetic sciences—operations research, systems
theory, game theory, and information theory—presented themselves with
seemingly cosmological force, delivering balance to a postwar world riven
by rage.
Modern computing talk owes a fair amount to these cybernetic sciences
as well. A visible contribution of cybernetics may be its consolidation and
popularization of a robust vocabulary for computing, including words such
as information, control, and feedback. In modern parlance, cybernetics also
gave currency to the widely used and now slightly pejorative prefix cyber
(-bully, -café, -crime, -dating, -fraud, -law, -punk, -security, -sex, -space, -ter-
rorism, -warfare) as well as the phrase “in the loop.” In popular culture,
cybernetics also helped breathe life into the scientific fictional imagina-
tion of the cyborg—or cybernetic organism—as an ensemble of human and
machinic parts, even though in practice formal cybernetics research rarely
dealt with cyborg research.3
For the purposes of this project, cybernetics sets the scene and props up
the intellectual scaffolding that is helpful for understanding the promises
and problems of cold war computing initiatives and sciences, including
the U.S. ARPANET, the Soviet OGAS, and OGAS’s sibling network projects.
In this chapter, I trace a brief global history of cybernetics, including its
sources and consolidation in postwar America, its spread to other cold
war climes and countries, and its adoption in post-Stalinist Soviet science.

9800.indb 16 6/2/16 3:05 PM

A Global History of Cybernetics  17

Backlit by observations about how cybernetics, like most scientific dis-

course, expresses itself in variable international dialects as well as common
metaphors (such as the human mind as a model information system for
designing other systems), I then detail four major stages in the history of
Soviet cybernetics in general and the rise of a peculiarly Soviet field—eco-
nomic cybernetics—on which subsequent chapters build.

The American Consolidation of Cybernetics

Norbert Wiener, the MIT mathematician, inveterate polymath, and son of

the founder of Slavic studies in America, is often credited with launching
cybernetics with his 1948 book Cybernetics, or Control and Communication
in the Animal and the Machine.4 How much of any scientific event can be
credited to one person is arguable, although we can at least credit Wiener
for helping to consolidate and coin under one label a series of intellec-
tual influences and sources. These sources were so complex and varied that
perhaps his greatest accomplishment was not setting into motion a new
field but synthesizing ideas from philosophy, mathematics, engineering,
biology, and literary and social criticism in his masterwork. Wiener’s input
exceeded even his output, which was tremendous. During World War II,
Wiener researched ways to integrate human gunner and analog computer
agency in antiaircraft artillery fire-control systems, vaulting his wartime
research on the feedback processes among humans and machines into a
general science of communication and control, with the gun and gunner
ensemble (the man and the antiaircraft gun cockpit) as the original image
of the cyborg.5
To designate this new science of control and feedback mechanisms,
Wiener coined the neologism cybernetics from the Greek word for steers-
man, which is a predecessor to the English term governor (there is a com-
mon consonant-vowel structure between cybern- and govern—k/g + vowel +
b/v + ern). Wiener’s popular masterworks ranged further still, commingling
complex mathematical analysis (especially noise and stochastic processes),
exposition on the promise and threat associated with automated informa-
tion technology, and various speculations of social, political, and religious
natures.6 For Wiener, cybernetics was a working out of the implications of
“the theory of messages” and the ways that information systems organized
life, the world, and the cosmos. He found parallel structures in the com-
munication and control systems operating in animal neural pathways, elec-
tromechanical circuits, and information flows in larger social systems.7 The
fact that his work speaks in general mathematical terms also sped his work’s

9800.indb 17 6/2/16 3:05 PM

18  Chapter 1

reception and eventual embrace by a wide range of readers, including Soviet

philosopher-critics, as examined later. Wiener placed little faith in his scien-
tific field to usher in peace—a social value disguised in his technical work
on homeostasis, a near synonym for dynamic equilibrium that he borrowed
from biology—into a world destabilized by mass violence. Nonetheless, his
thesis of the 1950 second edition of his masterwork Cybernetics prophesied
that “society can only be understood through a study of the messages and
the communication facilities which belong to it; and that in the future devel-
opment of these messages and communication facilities, messages between
man and machines, between machines and man, and between machine and
machine, are destined to play an ever-increasing part.”8
A second strand of American cybernetic thought, led by neurophysiolo-
gist Warren McCulloch, took seriously the brain-computer analogy—that
is, the now long-disputed notion that a brain can best be described as a
complex information processor, transmitter, and site of memory storage.9
McCulloch is remembered for his long white beard and contributions as the
organizer of the Macy Conferences on Cybernetics, which consolidated the
cybernetics movement in America. Researchers and historians of science
remember his 1943 paper, coauthored with the enigmatic polymath Wal-
ter Pitts, “A Logical Calculus of the Ideas Immanent in Nervous Activity,”
which proposed models for neural networks in the brain that later became
influential in the theory of automata, computation, and cybernetics. Their
argument holds that the mind is, given certain reductions, equivalent to a
Turing machine. In other words, with sufficient abstraction, it is possible
to imagine the neural network in a mind as a logical circuit that is capable
of carrying out any computable problem. In McCulloch’s words, he sought
“a theory in terms so general that the creations of God and men almost
exemplify it.”10
That “almost” packs much into its experimental epistemology. Although
the conclusion that the mind functions as a computer has since been dis-
puted and dismissed by several generations of neuroscience and cognitive
science, the basic neurophysiological insights that McCulloch brought
to cybernetics animated the midcentury cybernetic scene. These insights
included some inspiration for the development of distributed communica-
tion networking behind the ARPANET and to this day continue to inform
some contemporary artificial intelligence research. In what follows, I rein-
troduce his seminal but largely overlooked cybernetic notion of heterarchy
to understand dynamic networks of competing actors.
If cybernetics in the United States sprang from the teams of research-
ers channeling Wiener and McCulloch, it took disciplinary shape at the

9800.indb 18 6/2/16 3:05 PM

A Global History of Cybernetics  19

Macy Conferences on Cybernetics, a series of semiannual (1946–1947)

and then annual (1948–1953) interdisciplinary gatherings chaired by War-
ren McCulloch and organized by the Josiah Macy, Jr. Foundation in New
York City. The Macy Conferences, as they were informally known, staked
out a spacious interdisciplinary purview for cybernetic research.11 In addi-
tion to McCulloch, who directed the conferences, a few noted participants
included Wiener himself, the mathematician and game theorist John von
Neumann, leading anthropologist Margaret Mead and her then husband
Gregory Bateson, founding information theorist and engineer Claude Shan-
non, sociologist-statistician and communication theorist Paul Lazarsfeld,
psychologist and computer scientist J.C.R. Licklider, as well as influential
psychiatrists, psychoanalysts, and philosophers such as Kurt Lewin, F.S.C.
Northrop, Molly Harrower, and Lawrence Kubie, among others. Relying
on mathematical and formal definitions of communication, participants
rendered permeable the boundaries that distinguished humans, machines,
and animals as information systems. The language of cybernetic and infor-
matic analysis—including terms such as encoding, decoding, signal, feedback,
entropy, equilibrium, information, communication, control—sustained the anal-
ogies that bound together ontologically distinct physical phenomena.12
The “invisible college” constituted by the Macy Conferences proved
immensely influential:13 Von Neumann pioneered much of the digital
architecture for the computer as well as cold war game theory;14 Shannon
founded American information theory; Bateson facilitated the adaptation
of cybernetics in anthropology and the American counterculture;15 Lazars-
feld fashioned much of postwar American mass communication research;16
and of special note here, Licklider went on to pioneer and manage the U.S.
ARPANET (predecessor to the Internet) and its founding vision of human-
computer interaction. The effects of World War II on the global research
community shaped both the number of international participants in the
group (for example, von Neumann was a Hungarian émigré and Lazars-
feld was Viennese) as well as the distinctly American approach that the
Macy Conferences represented as a trading zone between private philan-
thropic institutions (the Macy Foundation) and academics with strong ties
to U.S. military research (including von Neumann, Wiener, Bateson, and
many others).17 Cybernetics emerged as a discipline that consolidated dis-
tinctly international sources of inspiration in a distinctly postwar American
The principles to emerge out of the Macy Conferences were many and far
from consensual. “Our consensus has never been unanimous,” McCulloch
quipped in a summary of the proceedings: “Even had it been so, I see no

9800.indb 19 6/2/16 3:05 PM

20  Chapter 1

reason why God should have agreed with us.” Nonetheless, a few remarks
help sketch out its conceptual pliability for later international interpreta-
tion. The first methodological hallmark of cybernetics is that it is not one
thing but that its key concepts, especially human-machine interaction and
feedback, outline a kind of vocabulary for working analogically across dif-
ferent systems—computational, mechanical, neurological, organic, social—
that rendered its vocabulary fecund for other sibling fields embedded in
U.S. military-industrial research.18
Take, for example, the contemporary fields of information theory and
game theory. Mainstream American information theory, following Bell
Labs engineer Claude E. Shannon’s 1948 mathematical theory of com-
munication, concentrates on the efficient and reliable measurement and
transmission of data.19 Perhaps its central seminal contribution is the theo-
rizing of a statistical framework for understanding all data transmissions.
All communication messages became a question of probabilities and sto-
chastic analysis, and the term information abandoned its ordinary meaning
of relevant facts and took on a new definition as a technical measure of the
likelihood that a message contains something ordered or surprising. Such
insights sped the theoretical development of computational communica-
tion systems, although Shannon’s theories were not widely applied until
the advent of affordable personal computers in the 1970s.
Von Neumann’s game theory (still influential in contemporary econom-
ics, business, and policy) developed formal models for human behavior
based on strategic and rational decision-making processes.20 By presuming
that the players in its games are rational actors seeking to make strategic
decisions, game theory formalized approaches to mathematically describ-
ing, modeling, and proscribing the optimal behaviors in both competitive
and cooperative multiplayer interactions that came to characterize the cold
war as a whole.21
The founders of these fields disagreed about the limits and relation-
ships between the three fields. Shannon insisted on keeping the technical
principles of information theory separate from the more sweeping scope
of cybernetics, Von Neumann did not rigorously distinguish between the
three, and Wiener defended his grouping of the other two research fields
under the cybernetics umbrella, even as (especially after mid-1950s) many
information theorists and game theorists objected to any conflation of
these fields.22 All three fields presented overlapping rational and general-
ized models of communication, or a “theory of messages” fit for applica-
tion, even though no one—not even the founders—knew the exact limits
of these computation communication sciences. Shannon did not accept

9800.indb 20 6/2/16 3:05 PM

A Global History of Cybernetics  21

the label of cybernetics, and he also did not accept the label others had
given to his own “information theory,” preferring to the end of his life
his original emphasis on “mathematical theory of communication.” Each
of these sciences sought to theorize the technical means by which com-
munication could be controlled. The cybernetic sciences, especially but not
exclusively in the Soviet case, emerge as a communication science in search
of self-governing systems.
Although it has never been clear (perhaps even to cyberneticists) what
cyberneticists could do exactly, it also never has been obvious what cyber-
netics could not do (perhaps even the definition of cybernetics is self-gov-
erning). For example, in 1943, Wiener and his coauthors succeeded in
springing “feed-back”—a once obscure term on loan from control engi-
neering and reclaimed in his antiaircraft research—into an umbrella con-
cept that was fit for understanding any type of purposeful behavior, where
the behavior of humans, animals, and machines is understood as “any
change of an entity with respect to its surroundings.”23 At this philosophi-
cal height, feedback loops proved a generalizable tool that could stabilize all
kinds of unsteady systems: feedback offers a process whereby information
that leaves a system is brought back into the system with the intention of
influencing that system’s future behavior.
Feedback comes in at least two kinds—positive and negative. When posi-
tive, feedback amplifies a signal cyclically, much like a microphone that is
set too close to a loudspeaker will cause painful audio feedback as the signal
loops out of control. Negative feedback by contrast can serve as a stabiliz-
ing agent, an internal check or correction on a system seeking balance in
an unstable environment. By working with feedback loops in communica-
tion systems, cybernetics sought a revolution in recognizing and operation-
alizing the nonlinear, self-recursive processes that abound in nature and
technology. Whatever cybernetics is, it is not a straightforward worldview
of Newtonian physics, Cartesian grids, Euclidian geometry, Aristotelian
cause and effect, and arithmetic. Rather, cybernetics espouses a mathemati-
cal worldview that helps us understand the midcentury struggle to balance
atop the tectonic shift in science toward pre-postmodern concepts such as
quantum physics, curvilinear grids, non-Euclidian geometry, cyclical cau-
salities, self-similar fractals, and modern probability theory.
The interpretive purchase of Wiener’s cybernetics rests not on its clarity
but on its synthetizing search for system self-regulation in the face of a topsy-
turvy postwar world. That is, the basic cybernetic approach seeks to har-
ness to the logical power of computing a wide range of scientific problems
with circularities and feedback loops. In this search for a balance between

9800.indb 21 6/2/16 3:05 PM

22  Chapter 1

the incongruities of material behavior and the sharp logic of computing,

feedback—even more than the cybernetic watchwords information, control,
and equilibrium—emerges as a clean concept for attempting to domesticate
all kinds of unruly communication systems. In 1943, Warren McCulloch
introduced the companion, although largely neglected, notion of heterarchy,
which serves as a useful lens for focusing scrutiny on the Soviet case. This
cybernetic concept helps describe some of the sources of conflict that beset
Soviet cybernetic attempts to network their command economy.
Let us set up this argument with a glimpse into institutional networks of
actors that are based on neither flat market nor hierarchical states but on
this third or middle way of heterarchy. The cold war organizational tropes
for self-regulation break down along that spectrum of economic order that
conventionally opposes market and hierarchy. In this view, the market is
understood as a flattened space for free interaction and efficient possibility
discovery among varied economic actors, and the hierarchy is understood
as a well-ordered, top-down pyramid of superiors over subordinates that is
well suited for completing long-term and complex tasks. Etymologically,
the English market is by far the newcomer of the two and can be traced back
to the mid-thirteenth-century Italian term for a “public building or space
for trading, buying, and selling.” The term market economy is first noted in
English only in 1948, centuries after the early modern capitalist revolution
that gave it fame and that has since enjoyed a privileged if often misunder-
stood position in the Western vocabulary of modern politics, economics,
and society. One reason for justifying the Pareto efficiency of the market
rests on the transitivity of human preferences. For the market to be the
ideal organizational mode, some economists assume that rational actors
will rank the order of their preferences linearly: if rational actors prefer
option A over B as well as option B over C, they also will prefer option A
over C. Yet this view of the market has been challenged in recent decades.
Markets hide transaction costs and information asymmetries. Behavioral
economists have demonstrated how under a number of conditions (such
as fear, regret, the threat of loss, cognitive dissonance, or peer pressure) the
rational homo economicus is a fiction: a person may prefer apples to bananas,
bananas to cantaloupes, and cantaloupes to apples, and there is no guaran-
tee that there exists a rational solution to voting systems or daily choices
involving three or more actors.24
By contrast, the concept of hierarchy (from the Greek term ἱεραρχία, “rule
by priests”) reaches back fifteen centuries to religious roots. As sociologist of
economics David Stark shows, the term was first used by a Christian medi-
eval theologian who is known today as the Pseudo-Dionysius the Areopagite.

9800.indb 22 6/2/16 3:05 PM

A Global History of Cybernetics  23

He published in the late fifth century under the pseudonym Dionysius the
Areopagite, a name first attributed to a first-century convert of Paul and
the first bishop of Athens. The fifth-century Christian mystic theologian
describes two far-reaching hierarchies in his Heavenly and Ecclesiastical Hier-
archies: the first, which includes nine levels of celestial beings (extending
from the supreme Godhead to the angels, who were just above humans),
serves as a symmetrical reflection for the second, which includes nine lev-
els of church leadership (seen in the current nine-tier Catholic ecclesiasti-
cal hierarchy descending from pope to bishop).25 The concept of hierarchy
has abounded in Western thought ever since—in the nine levels in Dante’s
Inferno; the organizational design of countless church, military, governmen-
tal organizations; and the conceptual imprint of information classification
systems, computer sciences, mathematics, and categorical thought. These
are all scalable approaches to bringing order in the modern world. Perhaps
the strongest example of hierarchy and socialism in modern America is also
its greatest bastion of patriotism—the U.S. armed forces, whose command-
and-control silos deliver social services and benefits to its members.
The logic of hierarchy has faced many challenges. Most modern critical
thought—epistemologists William James and Michel Foucault, critical the-
orists and feminists, Marxists and free-market theorists, liberation theorists
and theologians on the radical left and right, digital media theorists and
others—is organized against hierarchy.26 Even though the cold war ideo-
logical division over planned and free-market economies preoccupies fewer
social scientists today, modern organizational power and its resistance still
organize along the coordinates of hierarchy and open system.
The cybernetic concept of heterarchy offers a third way and an alternative
model between market and hierarchy that helps make sense of the Soviet
cyberneticists and informs later network analysis of how Soviet cyberneti-
cists tried to build computer networks to match the institutional networks
running the command economy. In 1945, just before McCulloch took
stewardship over the Macy Conferences, he published a five-page essay, “A
Heterarchy of Values Determined by the Topology of Nervous Nets,” in the
Bulletin of Mathematical Biophysics that coined the term heterarchy and estab-
lished how even the simplest systems can be subject to multiple competing
regimes of evaluation.27
Heterarchies are neither ordered nor disordered but instead are ordered
complexly in ways that cannot be described linearly. McCulloch takes as
his simplest example a network of three neurons arranged into a hierarchy
of transitive connections from neuron A to neuron B and from neuron B to
neuron C in which there are no “diallels” or “cross-overs.” His description

9800.indb 23 6/2/16 3:05 PM

24  Chapter 1

references the hierarchy in “the sacerdotal structure of the Church” in

which “the many ends are ordered by the right of each to inhibit all infe-
riors.” He then contrasts a hierarchical network with an intransitive neural
network in which a crossover is introduced between neurons C and A. In
this case, to model the network one needs to “map the network not on
a plane, but on a three-dimensional Taurus (a donut-shaped topological
space).” Instead of imagining such a network arrangement as inferior or
inconsistent, he observes that “circularities in preference actually demon-
strate consistency of a higher order than had been dreamed of in our phi-
losophy. An organism possessed of this nervous system—six neurons—is
sufficiently endowed to be unpredictable from any theory founded on a
scale of values. It has a heterarchy of values, and is thus internectively too
rich to submit to a summum bonum.”28
With the concept of neural heterarchy, McCulloch introduces the multi-
dimensional possibilities for complex systems that cannot be mapped onto
two-dimensional logics of either flat markets or tall hierarchies. This con-
cept has since proven helpful in cybernetic-compatible research far beyond
brain research, including self-organization, feedback loops, automata the-
ory, and non-Turing and non-Euclidean computing for thinking about the
superabundance of actual complex networked relations and also about the
limits of traditional tools for accounting for these relations. As detailed
later, similar cybernetic notions introduced both the terms and the network
tools for describing and managing the heterarchical tensions at the heart of
the Soviet command economy.

Cybernetics beyond the Cold War Superpowers

Between 1948 and the mid-1950s, cybernetics also enjoyed reception and
development in a number of countries outside of the cold war superpower
axis. For the purposes of this section, I focus on the postwar reception of
cybernetics in England, France, and Chile, although the point of the sec-
tion supersedes comparative local or national histories. The conditions of
modern countries after World War II and during the cold war were ripe
for an umbrella science of self-governance. Many scientists worldwide were
rushing to find ways to stabilize and regulate the consequences of a torrent
of new and disruptive technologies—and cybernetics modeled a technical
mindset for how to grapple with and control the consequences of tech-
nology itself. The 1950s saw a dizzying number of potentially revolution-
ary technologies become popular—atomic and hydrogen bombs, nuclear
power plants, Sputnik, the double helix, passenger jets, dishwashers, polio

9800.indb 24 6/2/16 3:05 PM

A Global History of Cybernetics  25

vaccines, the lobotomy (invented in the 1930s), television, and transistor

radios—and other trends, such as rock & roll and suburban housing devel-
opments. The disruptive influences of modern science and technology con-
tinued to be felt in the 1960s as quarks, lasers, Apollo, nylon, Pampers, the
pill, LSD, napalm, DDT, mutually assured destruction, and the ARPANET
entered the world stage. The most disruptive and destructive of all was the
development of computers around the work of John von Neumann at the
Institute of Advanced Studies at Princeton to study and control the effects
of nuclear bombs.29 The technocratic promise of the computer seemed to
promise both delivery and destruction. If computers could help civilize the
terrible and awesome power of the atom bomb, thought the scientists of
the day, then perhaps it might help stabilize lesser disruptions of modern
science and technology. If not, what terrible consequences would follow?
Or as von Neumann asked, taking the pulse of the moment in 1955: “Can
We Survive Technology?”30
Von Neumann’s question especially animated those who were engaged
in the nuclear cold war. In postwar France, United Kingdom, and Chile, the
potential of the computer to civilize awesome powers generated a “tech-
nology” of cybernetic interest in the 1950s and 1960s that was sometimes
more disruptive than the atomic bomb that troubled von Neumann. It was
the human mind imagined as an embodied machine. Might cybernetics
and its heir in cognitive science, midcentury scientists wondered, crack the
human mind and in turn spark new insights into how that most creative of
technologies might be modeled elsewhere?31
In France, the intellectual contributions of cybernetics began with more
analogies to politics than to the parietal lobe. Cybernetics had an early start
and a long afterlife in postwar France for several reasons. The public debate
about cybernetics turned the science into a bit of a political football between
communist and anticommunist debates in postwar France; the local intel-
lectuals helped ascribe a long French intellectual tradition to cybernetics,
which softened its reception; and Norbert Wiener visited France repeatedly
and promoted his science vocabulary in person. The imprint of cybernetics
can still be seen in subsequent generations of French theorists.
These postwar happenings are described briefly below. In 1947, the year
before he published Cybernetics with the MIT Press, Wiener attended Szolem
Mandelbrot’s congress on harmonic analysis in Nancy, France, which
resulted in a French book contract for the book that, while initially resisted
by the MIT Press, sold a sensational 21,000 copies over three reprints in six
months after its release in 1948. Three years later, in 1951, at the invitation
of Benoit Mandelbrot, the founder of fractals and Szolem’s nephew, Wiener

9800.indb 25 6/2/16 3:05 PM

26  Chapter 1

returned to lecture at Collège de France. Between 1947 and 1952, a flurry

of press coverage and public controversy sprung up between two camps of
anticybernetic communists and anticommunist cyberneticists.32 (Jacques
Lacan, who served in the French army, may very well have been among the
anticommunists and early cyberneticists at the time.) These debates over
the future of the governance of the French state were fueled by a slow and
painful postwar recovery, with widespread poverty aiding popular com-
munist and pro-Soviet sentiment. However, after Paul Ramadier’s socialist
party voted to accept the American Marshall Plan during the international
Paris meeting in 1947, anticybernetic communists slowly fell out of public
favor and with it, the debate about cybernetics. Similar to the initial Soviet
rejection of cybernetics, the initial public reaction to cybernetics appears
less about its science than about its status as an American import.33
At the same time, the public and implicitly pro-American defense of
cybernetics in the French press also helped reclaim this foreign science
as the heir to a distinctly French intellectual tradition that included the
rational mind-body concerns of René Descartes’ and Denis Diderot’s ratio-
nal encyclopedia, the physicist André-Marie Ampère’s coining of the term
cybernétique as a political science of peaceful governance in 1834, and the
structural linguistics of Ferdinand Saussure. Although I currently know of
no obvious direct connection between French cybernetics and the Minitel
network that developed between 1980 and 1989, the situation nonetheless
points to a generative and transnational intellectual exchange about the
scientific self-governance of a nation. As recent interpreters have argued,
many leading lights of postmodern French theory trace some of their basic
insights to French postwar cybernetic sciences. These include Claude Lévi-
Strauss’s treatment of language as a technologically ordered series (after
meeting with Macy Conference attendee Roman Jakobson in Paris in 1950);
Jacques Lacan’s turning to mathematical concepts; Roland Barthes’s turn to
schematic accounts of communication; Gilles Deleuze’s abandonment of
meaning, with Claude Shannon’s information theory in hand; Felix Guat-
tari’s, Michel Foucault’s, and other French theorists’ experimentation with
terms such as encoding, decoding, information, and communication.34 Post-
modern French theory owes a deep debt to postwar information theory
and the cybernetic sciences.
In England, cybernetics took on a different character in the form of the
Ratio Club, a small but potent gathering of British cybernetic figures who
gathered regularly in the basement of the National Hospital for Nervous
Diseases in London from 1949 through 1955. Notable figures include the

9800.indb 26 6/2/16 3:05 PM

A Global History of Cybernetics  27

computing pioneer Alan Turing, his Bletchey Park colleague and cryptog-
rapher mathematician I. J. Good, neuropsychologist Donald MacKay, and
astrophysicist Tommy Gold. The historian of science Andrew Pickering
chronicles the lives and work of six active and largely forgotten Britons who
were preoccupied with what the brain does—neurologist W. Grey Walter,
psychiatrist W. Ross Ashby, anthropologist, psychiatrist, and Macy Confer-
ence attendee Gregory Bateson, radical antipsychiatrist R. D. Laing, psy-
chologist Gordon Pask, and management cyberneticist Stafford Beer (who
also features prominently in the Chilean cybernetic situation described
below). Interdisciplinary discussions ranged widely across themes such as
information theory, probability, pattern recognition, artifacts that act (such
as William Ross Ashby’s homeostat and W. Grey Walter’s robotic tortoises),
and philosophy. Among their guests over the years were at least two Ameri-
cans who later played roles in the development of the ARPANET—J.C.R.
Licklider and Warren McCulloch.35 According to Pickering, what each of
these pioneering cyberneticists held in common was an interest in the
brain as a machine that acts, not thinks—or communication systems that
perform, not cogitate.36 Cybernetics took root on its own terms in Britain—
not the postmodern theory of France but what Pickering calls the “non-
modern” performances of neurological structures.
Chile and to a lesser extend Argentina also experienced the influx of
cybernetic ideas that ended up framing the debate about national networks.
In 1959, as a graduate student at Harvard, the Chilean biologist Humberto
Maturana coauthored an important paper, “What the Frog’s Eye Tells the
Frog’s Brain” with lead author Jerome Lettvin, Warren McCulloch, and
Walter Pitts. In the early 1970s, Maturana and his student Francisco Valera
secured their part in what has been called the “second wave” of cybernet-
ics together with the editor of the Macy Conferences proceedings Heinz
von Forester and Gordon Pask, among others, with their contribution of
the idea of autopoiesis—a system that generates, maintains, and reproduces
itself (such as a biological cell). The idea found resonance with the work on
the Chilean socialist economy led by the British management cyberneticist
Stafford Beer during the political rule of Salvador Allende between 1971
and the military coup in 1973.
During this period, Project Cybersyn took place, and it was perhaps the
most prominent experiment in developing a national network intended
for managing the socialist economy. As historian of science Eden Medina
has recently revealed, the British management cyberneticist Stafford Beer
served as principal architect for the rapid design, development, and partial
deployment of this nationwide network of telex machines connected to

9800.indb 27 6/2/16 3:05 PM

28  Chapter 1

a central mainframe computer. Beer, working with then finance minister

and engineer Fernando Flores, imported and adapted his (British) emphasis
on the brain as a model for managing organizations as published in the
1972 book The Brain of the Firm.37 His overtly cybernetic idea of a viable
system—a system that is designed to survive by adapting to its changing
environment—took root in the design of the Project Cybersyn network and
was reflected the political ideals of Allende’s democratically elected social-
ism and the autonomy of the workers. Despite limited success in rerouting
goods during a 1972 strike of truck drivers, the Cybersyn network, including
its futuristic central operations room, were scrapped in the military coup of
General Augusto Pinochet in 1973. In this Chilean case and perhaps in the
larger Latin American scene, cybernetics dovetailed with a strong emphasis
on embodied philosophy of mind.38
Before turning to the Soviet reception and translation of cybernetics, let
us look briefly at the eastern European sources of the cybernetic tradition,
some of which precede its consolidation in U.S. military research and the
postwar Macy Conferences. The list of precybernetic promoters includes
several notable figures. Aleksandr Bogdanov—old Bolshevik revolutionary,
right-hand man to Vladimir Lenin, and philosopher—developed a whole-
sale theory that analogized between society and political economy, which
he published in 1913 as Tektology: A Universal Organizational Science, a proto-
cybernetics minus the mathematics, whose work Wiener may have seen in
translation in the 1920s or 1930s.39 Stefan Odobleja was a largely ignored
Romanian whose pre–World War II work prefaced cybernetic thought.40
John von Neumann, the architect of the modern computer, a founding
game theorist, and a Macy Conference participant, was a Hungarian émigré.
Szolem Mandelbrojt, a Jewish Polish scientist and uncle of fractal founder
Benoit Mandelbrot, organized Wiener’s collaboration on harmonic analy-
sis and Brownian motion in 1950 in Nancy, France. Roman Jakobson, the
aforementioned structural linguist, a collaborator in the Macy Conferences,
and a Russian émigré, held the chair in Slavic studies at Harvard founded
by Norbert Wiener’s father. And finally, Wiener’s own domineering and
brilliant father, Leo Wiener, was a self-made polymath, the preeminent
translator of Tolstoy into English in the twentieth-century, the founder of
Slavic studies in America, an émigré from a Belarusian shtetl, and like his
son, a humanist committed to uncovering methods for nearly universal
Although summarizing the intellectual and international sources for the
consolidation of cybernetics as a midcentury science for self-governing sys-
tems is beyond the scope of this project, the following statement is probably

9800.indb 28 6/2/16 3:05 PM

A Global History of Cybernetics  29

not too far of a stretch. In each of the case studies examined here—War-
ren McCulloch’s heterarchical neural networks, the French evolution of
information theoretic and its turn to postmodern theory, the British Ratio
Club’s emphasis on performative models and agents, the Chilean build-
ing of a socialist national economic network after a model of the nation
as an organized firm, and sundry competing eastern European forces—the
midcentury cybernetic sciences are expressed in the local dialects of an
intellectual milieu and share with cognitive science an impulse to think
with the model of the mind. To a different effect, cyberneticists have been
constructing system analogs to understanding the mind and using such
mind models as analogs for reenvisioning new social, technological, and
organic worlds. The fascination with the mind is not new to cybernetics.
The millennia-long preoccupation with the inner workings of the mind, as
one neuroscientist quipped, may be little more than our own brain’s con-
ceit about itself.42

Soviet Cybernetics

With the first Soviet test of the atomic bomb in 1949, the cold war conflict
between capitalism and socialism slipped into the nuclear age. Soviet sci-
entists, philosopher-critics, and journalists redoubled their search for real
threats, as well as exciting possibilities, in the rapidly developing sphere
of science and technology, including rumors about a new American field
called cybernetics. Between 1947 (the year Norbert Wiener coined the term
cybernetics at a Macy Conference in New York) and 1953 (the year after
Joseph Stalin died), the state of Stalinist science, having proven itself as
essential to winning the war, enjoyed a complicated improvement in social
status, better funding, and uneven intellectual autonomy.43 The Soviet
Union stood out as a state that was committed to groundbreaking science.44
At the same time, certain fields of science, especially genetics in the wake
of the Lysenko debates, experienced acute pressures and censorship.45 And
although cybernetics was not outright repressed during Stalin’s rule, it was
widely ridiculed in the press and did not flourish until after his death. The
remainder of this chapter shows that even though post-Stalinist cybernetics
seemed poised to remake the Soviet Union as an information society, the
history of Soviet cybernetics, especially during the period of its rehabilita-
tion and adoption, slouches in significant ways toward the normal patterns
of Soviet history. In four overlapping sections below, I show that Soviet sci-
entific discourse rejected, rehabilitated, adopted, and adapted cybernetics
for historically expedient and changing purposes.

9800.indb 29 6/2/16 3:05 PM

30  Chapter 1

The Stalinist Campaign against Cybernetics: A “Normal” Pseudo-Science

Not all was rosy at the start. Amid abundant American accolades follow-
ing the publication of Wiener’s Cybernetics, or Control and Communication
in Animal and Machine in 1948, the Soviet press poured on insults. In 1950,
at the same time that the American Saturday Review of Literature was trium-
phantly proclaiming that it was “impossible for anyone seriously interested
in our civilization to ignore [Wiener’s Cybernetics]. This is a ‘must’ book for
those in every branch of science,” the leading literary Soviet journal Liter-
aturnaya gazeta was calling Wiener one of those “charlatans and obscuran-
tists, whom capitalists substitute for genuine scientists.”46 In a 1950 article
titled after the computing machine developed by Howard Aiken, “Mark
III, a Calculator,” Soviet journalist Boris Agapov ridiculed the sensational-
ist American press for its exultations about the coming era of “thinking
machines,” styling Norbert Wiener as an unknown figure “except for the
fact that he is already old (although still brisk), very fleshy, and smokes
cigars.” Commenting on a Time magazine cover of a computer dressed in
a military uniform, Agapov continued, “it becomes immediately clear in
whose service is employed this ‘hero of the week,’ this sensational machine,
as well as all of science and technology in America!”47 After Agapov’s 1950
article, Wiener’s Cybernetics was officially removed from regular circulation
in Soviet research libraries; apparently only secret military libraries retained
copies into the early 1950s.48
In 1951, a public campaign in the Soviet Union called the computer
hype in the United States a “giant-scale campaign of mass delusion of
ordinary people.” The 1951 volume Against the Philosophical Henchmen of
American-English Imperialism categorized cybernetics as part of a worrying
fashion around “semantic idealism” and dubbed cyberneticists “semanti-
cists-cannibals” for their recursive logics, especially self-informing feedback
loops. In addition to American cyberneticist Norbert Wiener, the volume
identified those belonging to the group of “semantic obscurantists” as
including logician-pacifist Bertrand Russell, his Cambridge colleague Alfred
North Whitehead, and Vienna Circle logical positivist Rudolf Carnap. Posi-
tivism, semiotics, and mathematical logic all appeared guilty of the cardi-
nal cognitivist belief that “thinking was nothing else than operations with
signs.”49 In 1952, Literaturnaya gazeta ran an article called “Cybernetics: A
‘Science’ of Obscurantists,” which cleared the way for a deluge of popular
titles: “Cybernetics: An American Pseudo-Science,” “The Science of Modern
Slaveholders,” “Cybernetics: A Pseudo-Science of Machines, Animals, Men
and Society,” and so on.50

9800.indb 30 6/2/16 3:05 PM

A Global History of Cybernetics  31

In 1953, an author who wrote under the pseudonym “Materialist” pub-

lished the infamous article “Whom Does Cybernetics Serve?” in a lead-
ing journal for ideological and intellectual battles, Questions of Philosophy.
“Materialist” waxes poetic in his rebuke:

the theory of cybernetics, trying to extend the principles of modern computing ma-
chines to a variety of natural and social phenomena without due regard for their
qualitative peculiarities, is mechanicism turning into idealism. It is a sterile flower of
the tree of knowledge arriving as a result of a one-sided and exaggerated blowing up
of a particular trait of epistemology.51

Later in the article, Materialist contends that “in the depth of their
despair, [those in the capitalist world] resort to the help of pseudo-sciences
giving them some shadow of expectation to lengthen their survival.”52
With somewhat less vitriol, in 1954, the fourth edition of the Concise Dic-
tionary of Philosophy cast cybernetics as a slightly ridiculous, although still
harmful anti-Marxist “reactionary pseudo-science.” The entry reads:

Cybernetics: a reactionary pseudo-science that appeared in the U.S.A. after World

War II and also spread through other capitalist countries. Cybernetics clearly reflects
one of the basic features of the bourgeois worldview—its inhumanity, striving to
transform workers into an extension of the machine, into a tool of production, and
an instrument of war. At the same time, for cybernetics an imperialistic utopia is
characteristic—replacing living, thinking man, fighting for his interests, by a ma-
chine, both in industry and in war. The instigators of a new world war use cybernet-
ics in their dirty, practical affairs.53

The campaign continued in the popular and scholarly press more or less
unabated through the 1950s, although the first public rehabilitation efforts,
noted below, began in earnest as early as 1955.
The list of epithets reserved for cybernetics by the Soviet press should
be put into perspective. The campaign against cybernetics, however mean-
spirited and aggressive, appears far from the most vicious of campaigns
that were organized by Soviet journalists and public commentators against
American thought. Stalin, who was known to read widely across the sci-
entific fields, seems to have known little to nothing about cybernetics; his
fury against it appeared independent of “any essential features of cybernet-
ics itself,” according to Gerovitch.54 Without any direct evidence of Sta-
lin’s involvement in the campaign against cybernetics, we can speculate
that Stalin likely reviled cybernetics for the same reasons that he hated
all imperialist “pseudo-sciences”: ideological opposition was necessary to
fuel and power his monumental state building and modernization proj-
ects. The campaign against cybernetics, which came in the wake of Stalin’s

9800.indb 31 6/2/16 3:05 PM

32  Chapter 1

personal affront against classical genetics, appeared more or less a “farce”

to some philosopher-critics. These same philosopher-critics, according to
information theorist Ilia Novik, “berated cybernetics with certain … indif-
ference and even fatigue.” In the late 1940s and early 1950s, as cybernetics
was sweeping the United States, France, the United Kingdom, Chile, and
other countries with the enthralling possibilities of self-organizing human-
machine ensembles and predictive negative feedback loops, “cybernetics”
in the Soviet Union had, to crib Novik’s phrase, “emerged as a normal
The anti-American Soviet campaign against cybernetics was only one
among a range of operations that were meant to repress the Soviet knowl-
edge base, including but not limited to Stalinist science. A few other exam-
ples include the rise of Trofim Lysenko in Soviet biology, whose program
on the heritability of acquired characteristics ousted the study of Men-
deleev and classical genetics; the condemnation of Linus Pauling’s struc-
tural resonance theory by Soviet chemists in 1951; the banning of Soviet
Lev Vygotsky’s work, now recognized as a foundation of cultural-historical
psychology; the forestalling of structural linguistics pioneered by Ferdi-
nand Saussure, Nikolai Trubetzkoi, and Roman Jakobson; and the excoria-
tion of Albert Einstein’s theories of general and special relativity, quantum
mechanics, and Werner Heisenberg’s principles of indeterminacy as distor-
tions and corruptions of the true (that is, Marxist) objective and material
nature of the universe.56 In light of these and other examples, the public
campaigns against cybernetics strike the contemporary observer as far from
masterfully orchestrated or even normal in their regularity. The ground
warfare of ideological critique was messy, full of ritual elements, political
posturing, and routine debates. Not only did the enterprise of Soviet cyber-
netics prove to be diverse, but the anticybernetic campaigns that preceded
it varied richly.
There was nothing particularly anticybernetic about the early anticy-
bernetic campaigns. Rather, the early opposition to the science appears
overwhelmingly anti-American in motivation. In the decade that followed,
Soviet cybernetics transformed into an apparent harbinger of social reform
and later into a normal Soviet science. Even the Soviet ideological resis-
tance to cybernetics appears normal from the beginning.

The Post-Stalinist Rehabilitation of Cybernetics, 1954 to 1959

Natural Science will in time incorporate into itself the science of man, just as the
science of man will incorporate into itself natural science: there will be one science.
—Karl Marx, Economic and Philosophic Manuscripts of 1844

9800.indb 32 6/2/16 3:05 PM

A Global History of Cybernetics  33

Stalin’s death in March 1953 made possible a watershed shift in public dis-
course in favor of Soviet cybernetics and gave root to the promise of cyber-
netic-led structural reform of the Soviet Union—and especially the promise
of a new kind of self-governance in the wake of Stalin’s bloody rule. After he
seized power from his rivals in 1955, Nikita Khrushchev titled himself first
secretary, not general secretary as Stalin had, in an effort to signal a clean
break from the past and the launching of a new post-Stalinist era. Typically,
the only thing remembered about the Twentieth Congress of the Commu-
nist Party of the Soviet Union, is Khrushchev’s 1956 “secret speech,” which
he delivered to a carefully selected crowd and in which he became the first
Soviet authority figure to denounce Stalin’s crimes and the now infamous
“cult of personality.” The speech inaugurated the Khrushchev thaw, a period
known for the easing of censorship and political repression and the partial
de-Stalinization of Soviet policy, international relations, and society. These
public revelations, combined with a sagging Soviet economy, compelled even
those least likely to decry the terrible reality of Stalin’s terror to admit that, in
Khrushchev’s terms, “serious excesses” and “abuses” had been committed.57
As part of this sweeping technical reform, the new first secretary also
called for an ideological reappraisal of Marxism-Leninism:

In this connection we will be forced to do much work in order to examine criti-

cally from the Marxist-Leninist viewpoint and to correct the widespread erroneous
views connected with the cult of personality in the sphere of history, philosophy,
economy, and other sciences, as well as in literature and the fine arts. It is especially
necessary that in the immediate future we compile a serious textbook of the history
of our Party, which will be edited with scientific Marxist objectivity.58

By 1959, Stalin’s Short History of the Communist Party of the Soviet Union,
once characterized as “the catechism of Communism,” had been officially
deemed full of errors and withdrawn under Khrushchev. It was replaced in
1961 by the 900-page Fundamentals of Marxism-Leninism.59
The time between Stalin’s death and cybernetics’ entrance into the favor
of the press and Soviet public discourse on science was not great. In fact, in
the same 1956 Congress that he gave his “secret speech,” Khrushchev also
promoted cybernetic-friendly principles for automating the Soviet econ-
omy: “The automation of machines and operations,” he declared, “must be
extended to the automation of factory departments and technological pro-
cesses and to the construction of fully automatic plans.”60 With the pass-
ing of Stalin, cybernetics entered Soviet technical, scientific, and political
discourse at a time that was particularly primed for reform.
Although Soviet science enjoyed reforms and looser ideological con-
straints under Khrushchev, Soviet science may have accomplished more

9800.indb 33 6/2/16 3:05 PM

34  Chapter 1

under the fist of Stalin than it did under the loose umbrella of cybernet-
ics. Under Stalin, Soviet physicists and chemists pioneered work for which
chemist Nikolai Semyonov, physicist Igor Tamm, economist Leonid Kan-
torovich, and physicist Pyotr Kapitza received Nobel Prizes decades later.
Other Soviet scientists—including Igor Kurchatov, Lev Landau, Yakov Fren-
kel, and Andrei Sakharov, and other world-renowned figures—also devel-
oped atomic and thermonuclear bombs, a lynchpin in Stalin’s rapid and
forceful industrialization of the remnants of the Russian empire from a
backwater country into a global super power in the period of a few decades.
Many Soviet scientists successfully employed dialectical materialism as a
genuine source of inspiration, not a forced ideology, in their scientific work.
The reality that the health of science depended more on funding than it did
on freedom also sobers reflection on the contemporary state of science and
public attitudes about it.61
Soviet cybernetics arrived at a time that was well suited for leveraging
a post-Stalinist revision of scientific Marxist objectivity. It introduced its
mind-machine analogies in a light that was friendly to Ivan Pavlov’s cel-
ebrated notion of “conditioned reflexes” in psychology, which were based
on the reflex-response analogy of a telephone electrical switchboard, the
reactions of which depended on the programmable configuration of wires.
Both Pavlov and, two generations later, cyberneticists worldwide imagined
the mind as neural networks and electronic processors, a seminal metaphor
for what philosopher Pierre Dupuy dubbed the “mechanization of mind”
powering the subsequent rise of cognitive science.62
Soviet cybernetics also found the support of several world-famous math-
ematicians, which was a field in which the Soviets were internationally
recognized. Figures including Andrei Kolmogorov, Sergei Sobolev, Aleksei
Lyapunov, and Andrei Markov Jr., came together, despite significant differ-
ences, to form an early core of Soviet cybernetic mathematicians who were
committed to advancing this new metamathematical science as a single
science for Soviet thought. And just as cybernetics was mobilizing its intel-
lectual defenses, it also found institutional fortification in the creation of
Akademgorodok, a new “scientific township at Novosibirsk” in Siberia. Cre-
ated in the spring of 1957, this city of science (formally part of the city of
Novosibirsk) proved a refuge of privilege and relative intellectual freedom
for over 65,000 Soviet scientists, including Aleksei Lyapunov, a pioneering
Before the Soviet scientific mainstream could adopt cybernetics, the
attendant scholarly communities had to be prepared for an about-face in
the official Soviet attitude toward an American-born discipline. The first

9800.indb 34 6/2/16 3:05 PM

A Global History of Cybernetics  35

sign of this turnaround came not from Moscow but from a neighbor in
the near abroad: in 1954 in Warsaw, six “Dialogues on Cybernetics” sur-
faced, and they approached cybernetics in a critical dialectical tone that
was serious enough to signify that the topic deserved real discussion.64 In
the meantime, three mathematicians and an unlikely philosopher-critic
closer to Moscow set off on a mission to remake Soviet cybernetics from
the inside out.

The First Soviet Cyberneticists: Kitov, Lyapunov, Sobolev

In 1955, two Russian-language articles appeared in the same issue of the
Soviet journal Voprosi Filosophii (Problems of Philosophy), where “Materialist”
and others had railed against cybernetics in 1953. This signaled a watershed
change in the official attitude toward cybernetics. A closer look at these two
articles sheds light on this reversal. Sergei Sobolev, Aleksei Lyapunov, and
Anatoly Kitov coauthored the article titled “The Main Features of Cyber-
netics” and began the process of rehabilitating cybernetics from positions
of relative authority in the Moscow military-academy complex. Although
Kitov was the youngest and the least influential of the three mathematician
coauthors, he also appears to have been the first Soviet cyberneticist.
A Soviet colonel engineer, Anatoly Kitov discovered in 1952 the single
copy of Wiener’s Cybernetics in a secret library of the Special Construc-
tion Bureau—SKB-245—at the Ministry of Machine and Instrument Build-
ing. Kitov had been sent there to research possible military applications
for computers after graduating in 1950 from the military academy where
Lyapunov taught with a gold medal, the highest award in the Soviet educa-
tion system. After reading Wiener’s Cybernetics, Kitov began to consider that
cybernetics was, in his words, “not a bourgeois pseudo-science, as official
publications considered it at the time, but the opposite—a serious, impor-
tant science.”65
After digesting Cybernetics, Kitov turned to share his newfound enthusi-
asm for the science with his former instructor, Aleksei Lyapunov. Lyapunov,
who later was known as “the father of Soviet cybernetics,” was a wide-
ranging and luminous mathematician who taught at the Military Artillery
Engineering Academy and in the department of computational mathemat-
ics at Moscow University. Recognized by biologists, geophysicists, and phi-
losophers alike, Lyapunov took, according to Soviet historian of science
M. G. Haase-Rapoport, an “integrating, non-dividing approach in natural
science,” which “became the rich soil [for] the sprout of cybernetic ideas.”66
Having heard his case, Lyapunov in turn encouraged Kitov to write an arti-
cle explaining the essence of cybernetics, promising to coauthor it with

9800.indb 35 6/2/16 3:05 PM

36  Chapter 1

him. Holed up in the secret military research library, Kitov wrote a draft
for the article, after which Lyapunov recommended inviting as coauthor
Sergei Sobolev, then chair of the department of computational mathemat-
ics at Moscow University. Sobolev also played a legitimizing role as deputy
director of the Institute of Atomic Energy, in effect the mathematician with
a hand on the atomic bomb. In 1933, at the age of twenty-five, Sobolev
became the youngest corresponding member of the Soviet Academy of Sci-
ences, and in 1939, the youngest full member (academician) of the Acad-
emy. After joining the Bolshevik Party in 1940, Sobolev was appointed as
the deputy director of the Institute of Atomic Energy in 1943 and contrib-
uted to the construction of the first Soviet atomic and hydrogen bombs.
With this in mind, Lyapunov and Kitov arranged to visit Sobolev at his
dacha in Zvenigorod, an hour west of Moscow, where, after discussing
the draft, Sobolev offered his name as coauthor. Although it is not known
how much he contributed to the article, Sobolev repeatedly and publicly
defended cybernetics in the late 1950s.67
Sometime in 1952, Kitov and Lyapunov visited the editorial staff of Prob-
lems of Philosophy. For unknown reasons, the editors agreed to publish the
article, asking only that they receive permission from the Communist Party
first. We may speculate on why the editors agreed to publish on a forbid-
den topic. Voprosi Filosofii continued to publish anticybernetic material
for several years, so one might suppose that the editors thought permis-
sion would not be granted, thus shifting the blame for the rejection onto
higher authorities. It is equally possible that the editors agreed to publish
the article out of genuine enthusiasm to encourage intellectual debate dur-
ing Khrushchev’s thaw. Regardless, the editors sent Lyapunov and Kitov
to meet with representatives in the science division of Staraya Square, an
administrative wing for the Communist Party in downtown Moscow. The
administrators heard their case, asked some questions, and then concluded:
“We understand: it is necessary to change the relationship to cybernetics,
but an instantaneous split is not possible: before the article can be pub-
lished, it would make sense to do several public reports.”68 Lyapunov and
Kitov spent 1953 and 1954 carrying out tacitly approved public lectures
and private workshops, and Lyapunov began hosting in his home a circle
of colleagues to discuss cybernetics that lasted over a decade.69
At once an introduction, a reclamation, and a creative translation of Wie-
ner’s Cybernetics, Kitov, Lyapunov, and Sobolev’s feature article, “The Main
Features of Cybernetics,” danced a deliberate two-step. First, it attempted
to upgrade cybernetics to parity with other natural sciences by basing an
ambitiously comprehensive theory of control and communication almost

9800.indb 36 6/2/16 3:05 PM

A Global History of Cybernetics  37

exclusively on Wiener’s 1948 book (although these early Soviet cybernet-

ics made notably less of the field as an applied science and more of it as a
universalizing theory than did Wiener). Second, it retooled Wiener’s con-
ceptual vocabulary into a Soviet language of science. Gerovitch details the
translation of their terms: “What Wiener called ‘the feedback mechanism’
they called ‘the theory of feedback’ … ‘basic principles of digital comput-
ing’ became ‘the theory of automatic high-speed electronic calculating
machines’; ‘cybernetic models of human thinking’ became the ‘theory of
self-organizing logical processes.’”70 In fact, the coauthors used the word
theory six times in their definition of cybernetics to emphasize the theoreti-
cal nature of the new science, possibly as a way to avoid having to discuss
the political implications of introducing a practical field of human-machine
applications into a society well suited to adopt them.
The coauthors also integrated and expanded the stochastic analysis of
Claude Shannon’s information theory while simultaneously stripping Wie-
ner’s organism-machine analogy of its political potency.71 Wiener’s core
analogies between animal and machine, machine and mind were stressed
as analogies—or how “self-organizing logical processes [appeared] similar
to the processes of human thought” but were not synonyms. At the same
time, the article scripts his language of control, feedback, and automated
systems in the machine and organism into the common language of infor-
mation, or Shannon’s mathematical theory of communication. For Kitov,
this “doctrine of information” took on wholesale the task of universaliz-
ing statistical control in machines and minds. It did so by preferring the
“automatic high-speed electronic calculating machine” (that is, computer)
to Wiener’s original base analogy for cybernetic comparisons—the servo-
mechanism. The servomechanism is an automatic engineering device used
in a larger mechanism to correct, using error-sensing negative feedback,
that mechanism’s performance: examples could include the steam engine
governor, modern cruise control in cars, or, in Wiener’s case, antiaircraft fire
control mechanisms controlling a gun and its gunner.72 Despite the coau-
thors’ efforts to silence the social implications of the theory, computer algo-
rithms added a further layer of technical complication to Wiener’s feedback
mechanisms, even as their neuronal analog to electronic switches quietly
implied opening new research horizons in human-computer interaction,
robotic prosthetics, and cyborgs. By formulating the science in terms of
cutting-edge computers, not servomechanisms, the coauthors propelled
the Soviet cyberneticist and his computer into the front lines of the esca-
lating space and technology race. Thus, conceiving of the computer as a
general regulating machine for any control systems, the Soviet formulation

9800.indb 37 6/2/16 3:05 PM

38  Chapter 1

of cybernetics focused on computational systems from the start—a general-

ized step away from Wiener’s interests in communication and control in
concrete entities of “the animal and the machine.”73 Although computers
were not common in the Soviet Union until decades later, to this day, the
Russian word for cybernetics, kibernetika—together with its heir informatics,
or informatika—remains a near synonym for the English field of computer
Computers at the time were new media in the sense that few people
agreed how to talk about them: the computer in Russian in the 1950s and
1960s went by the bulky description “automatic high-speed electronic cal-
culating machine.”74 Frequent use of the term mercifully introduced the
abbreviation EVM (electronnaya vyichislitel’naya mashchina, or “electronic
calculating machine”), which stuck through the 1960s and 1970s. Only
under Gorbachev’s perestroika in the 1980s did the now nearly ubiquitous
English calque komp’yuter replace the term EVM.75 The unwieldiness of the
original Soviet term underscores the perennially renewable nature of the
discursive contest that makes computers more or less new. Because the
coauthors were sensitive to how language, especially foreign terms, packs
in questions of international competition, the coauthors attempted to keep
their language as technical and abstract as possible, reminding the reader
that the cybernetic mind-machine analogy was central to the emerging sci-
ence but should be understood only “from a functional point of view,” not
a philosophical one.76
The technical and abstract mathematical language of Wiener’s cybernet-
ics thus served as a political defense against Soviet philosopher-critics and
as ballast for generalizing the coauthors’ ambitions for scientists in other
fields. They employed a full toolbox of cybernetic terminology, including
signal words such as homeostasis, feedback, entropy, reflex, and the binary
digit. They also repeated Wiener and Shannon’s emphases on probabilistic,
stochastic processes as the preferred mathematical medium for scripting
behavioral patterns onto abstract logical systems, including a whole section
that elaborated on the mind-machine analogy with special emphasis on
the central processor as capable of memory, responsiveness, and learning.77
Wiener’s call for cyberneticists with “Leibnizian catholicity” of scientific
interests was tempered into its negative form—a warning against disciplin-
ary isolationism.78
On the last page of the article, the coauthors smoothed over the adop-
tion of Wiener, an American, as foreign founder of Soviet cybernetics
by summarizing and stylizing Wiener’s “sharp critique of capitalist soci-
ety,” his pseudo-Marxist prediction of a “new industrial revolution” that

9800.indb 38 6/2/16 3:05 PM

A Global History of Cybernetics  39

would arise out of the “chaotic conditions of the capitalist market,” and
his widely publicized postwar fear of “the replacement of common work-
ers with mechanical robots.”79 A word play in Russian animates this last
phrase: the Russian word for worker, or rabotnik, differs only by a vowel
transformation from robot, the nearly universal term coined in 1927 by the
playwright Karel Capek from the Czech word for “forced labor.”80 The first
industrial revolution replaced the hand with the machine, or the rabotnik
with the robot, and Wiener’s science, the coauthors dreamed, would help
usher in a “second industrial revolution” in which the labor of the human
mind could be carried out by intelligent machines, thus freeing, as Marx
had intimated a century earlier, the mind to higher pursuits. “Automation
in the socialist society,” the coauthors wrote in anticipation of Khrush-
chev’s declaration at the 1956 Congress, “will help facilitate and increase
the productivity of human labor.”81 Although Stalin had found no use for
Wiener’s sounding of a “new industrial revolution,” these mathematicians
had found and refashioned in Wiener an American critic of capitalism, a
founder of a science that was fit to sound the Soviet call for the “increased
productivity of labor.”82
Given this explicit adoption of Wiener into the Soviet scientific canon,
it is surprising to note that the coauthors quoted only one line from any of
his works. That line reads: “Information is information, not matter and not
energy. Any materialism that cannot allow for this cannot exist in the pres-
ent.”83 By distinguishing between information, energy, and matter, Wiener
skips across two recent paradigm shifts in modern physics—first, from a
Newtonian physics of matter to an era of thermodynamics and Bergson and
second, from the thermodynamics of energy to a new but related paradigm
of information science and Wiener’s cybernetics. For many in the West, this
quote meant that information is nothing but information, a value-neutral
statistical measurement on which to rest objective science and the search
for computable truth. The technical meaning was the same for their Soviet
counterparts, but it also meant something more. By singling out Wiener’s
alliance of materialism and cybernetics, the coauthors implied that Wiener
had in mind a position that was amendable to the official philosophy of
Soviet science—the dialectical materialism of Marxism-Leninism. If dialec-
tical materialism did not update itself for the information age, it could not
exist. The same quote also leaves open the opportunity that the coauthors
were lobbying for—that Soviet dialectical materialism could allow for infor-
mation to be information in its fullest cybernetic or stochastic sense. The
quote thus renders Wiener as a sort of foreign prophet announcing a dia-
lectical materialist science of information science, a science whose present

9800.indb 39 6/2/16 3:05 PM

40  Chapter 1

materialism could only be fully Soviet. With these ritual words, the coau-
thors wed cybernetics to Soviet ideology and dialectical materialism to the
cybernetic information sciences. The success of this “important new field”
of Marxist-Leninist information science, they contended, hung on the call
to action that was voiced by its American originator.
The coauthors also buttressed Wiener’s ideas of neural processing with
reference to the great Soviet scientist Ivan Pavlov, whose original theory of
conditioned reflexes in human psychology was derived from a telephone
electrical switchboard, a communication machine with ideal cybernetic
resonance.84 Finally, the coauthors concluded the article in a ritual flour-
ish of Orwellian newspeak that was common to academic writing at the
time, calling for a battle against the capitalists who “strive to humiliate the
activity of the working masses that fight against capitalist exploitation. We
must decisively unmask this hostile ideology.”85 After years of anti-Ameri-
can, anticybernetic positions, they were the first to voice an anti-American,
procybernetic position in the Soviet press. In the mid-1950s, the tone of
subsequent arguments began to distinguish between the capitalist use of
cybernetics, which was flatly condemned, and cybernetics in general, thus
creating space for the argument that the socialist use of cybernetics might
not only be possible but even preferable.

“The Dark Angel”: Ernest Kolman’s “What Is Cybernetics?”

Whatever rhetorical flourishes Kitov, Lyapunov, and Sobolev mustered,
the strongest ideological support for their newfound procybernetic posi-
tion lay in the article that immediately followed their publication in the
same journal, Ernest Kolman’s piece “What Is Cybernetics?” (“Chto takoe
kibernetika?”). A loyal Bolshevik, an active ideologue-philosopher, and a
failed mathematician with a long and bloody personal history of attack-
ing nonorthodox mathematicians, Kolman makes a somewhat surprising
candidate for the first ideological defender of Soviet cybernetics.86 Among
other ideological offenses that he appears to have committed, he seems to
have done the most harm to the founders of the Moscow School of Math-
ematics, a powerful school in imperial Russia and the Soviet Union. He
excoriated them for their nonatheistic commitment to a fascinating intel-
lectual alliance between French set theory and a Russian Orthodox name-
worshipping mysticism. (Their scandalously religious observation began
by noting that both infinity and God could be named but not counted.)87
Kolman was once dubbed “one of the most savage Stalinists on the front of
science and technology” for his tireless defense of Lysenko’s biology (which
is now remembered as the Soviet pseudo-scientific alternative to classical

9800.indb 40 6/2/16 3:05 PM

A Global History of Cybernetics  41

genetics).88 Some Soviet commentators feel that Kolman’s diatribes kept

the mathematician Andrei Kolmogorov in the 1940s from beating Wie-
ner—the two are often compared as intellectual peers—to formalizing the
link between biology and mathematics. Kolman was sensitive to political
attacks and had a genuine interest in the history of science and a knowl-
edge of four or five languages. A formidable opponent, he was sometimes
known among his detractors and victims as the “dark angel.”89
Despite such a body count, Kolman’s role as self-elected guardian of
cybernetics was not the first time he had deviated from an ideologically
orthodox line of philosophy. He had spent time in a Stalinist labor camp
after World War II for straying from the party line in his interpretation of
Marxism. Just before he died in 1982, he published the book We Should
Not Have Lived That Way, in which he reflected on his own past transgres-
sions: “In my time I evaluated many things, including the most important
facts, extremely incorrectly. Sincerely deluded, I was nourished by illusions
which later deceived me, but at that time I struggled for their realization,
sacrificing everyone.”90 This context makes Kolman’s defense of cybernet-
ics more surprising: why would an embittered former mathematician with
a track record of decimating pseudo-scientific mathematical theories come
to the defense of cybernetics in 1953? Was his role as the first ideologue to
defend Soviet cybernetics an act of penitence or another cardinal sin?
Kolman began his eleven-page promotional history by outlining over a
century of international cybernetics, beginning with the French mathema-
tician, physicist, and philosopher André-Marie Ampère in 1843 and moving
to “Russian and Soviet scientists, [such as] Chernishwev, Shorin, Andropov,
Kulebakin, and others.”91 Kolman called Wiener “one of the most visible
American mathematicians and professor of mathematics at Columbia Uni-
versity” and the one who “definitively” formalized cybernetics “as a scien-
tific sphere,” in a veritable shout of praise for the time.92 In fact, Wiener
had been appointed at MIT, not Columbia, since 1919, but Kolman may
have introduced the mistake on purpose: Columbia University stood out
to Soviet observers among American universities at the time for its Russian
studies center, the Harriman Institute, which had been a favorite target
of McCarthy, so by connecting Wiener to Columbia, not MIT, perhaps he
softened his image in the eyes of Kolman’s peer philosopher-critics.93 In any
case, Wiener occupies the sixth through the ninth paragraphs of Kolman’s
ideological support piece, which signals a second witness of Wiener’s adop-
tion into the vanguard of Soviet cybernetic historiography.
Having set up Wiener as the foreign founder of Soviet cybernetics in the
article, Kolman promptly invented a Soviet prehistory to the science that

9800.indb 41 6/2/16 3:05 PM

42  Chapter 1

broadened and colored the ambition of cybernetics to match Marxism-

Leninism. Sensitive to the many eastern European origins of cybernetic-style
thinking, Kolman’s narrative assimilates cybernetics into a longer history of
computational machines, including Ramon Llull in 1235, Pascal in the mid-
1600s, the engineer Wilgott “Odhner of St. Petersburg” (and not Stockholm,
Wilgott’s native city), and the late nineteenth-century mathematicians A.
N. Krilov and P. L. Chebishev. He then discussed how the Soviet mathemati-
cians Andrei Markov Jr. (a constructivist mathematician who later became
a leading cyberneticist), N. C. Novikov, N. A. Shanin, and others advanced
the last hundred years’ worth of precybernetic work in Russian.94 Kolman’s
internationalism allowed two people west of Berlin to slip into his history—
Norbert Wiener and Nikolai Rashevsky, the first Pavlov-inspired biomath-
ematician and a Russian émigré at the University of Chicago.
Thus, the battle to legitimize Soviet cybernetics began internally and was
fought against by and among Soviet philosopher-critics, the vanguard and
police of ideological debate in Soviet discourse. Both procybernetic articles
(especially Kolman’s) were loaded with discursive tactics that were meant
to protect cybernetics from counterattacks, so much so that, even in pro-
nouncing it, the first act of Soviet cybernetics partook in cold war game-
theoretic strategies. In the first public defense of cybernetics, which was a
lecture given at Moscow State University in 1954, Kolman notes that “it is,
of course, very easy and simple to defame cybernetics as mystifying and
unscientific. In my opinion, however, it would be a mistake to assume that
our enemies are busy with nonsensical things, that they waste enormous
means, create institutes, arrange national conferences and international
congresses, publish magazines—and all this only for the purpose of dis-
crediting the teachings of Pavlov and dragging idealism and metaphysics
into psychology and sociology.” By imagining enemies as rational actors,
not pseudoscientific bourgeois, a cybernetic worldview provides its own
first defense: “There are more effective and less expensive means than the
occupation with cybernetics,” Kolman the philosopher-critic continues, “if
one intends to pursue idealistic and military propaganda.95
Kolman employed the logic of reversing the rational enemy that was
implicit in all Soviet cybernetic strategy to save the fledgling movement
from future Soviet critics. Kolman invites his Soviet listeners to consider
cybernetics from the perspective of an economically rational American sci-
entist.96 We should imitate the enemy, Kolman reasoned, because we can
infer that the enemy knows something we do not, for he is occupied with
something we do not understand. To its participants, cybernetics took ini-
tial shape in a militarized discourse of the postwar and cold war.

9800.indb 42 6/2/16 3:05 PM

A Global History of Cybernetics  43

Like Kolman, the coauthors Sobolev, Lyapunov, and Kitov also pre-
empted the reactions of the Soviet philosophers, rebuffing them for “mis-
interpreting cybernetics, suppressing cybernetic works, and ignoring the
practical achievements in this field.”97 The coauthors flipped the reaction-
ary argument that was sure to follow (that Soviet cybernetic defenders were
“‘kowtowing’ before the West”) by insisting that “some of our philosophers
have made a serious mistake: without understanding the issue, they began
by denying the validity of a new scientific trend largely because of the sen-
sational noise made about it abroad.”98 In a concluding flourish, the coau-
thors conspired:

One cannot exclude the possibility that the hardened reactionary and idealistic in-
terpretation of cybernetics in the popular reactionary literature was especially orga-
nized to disorient Soviet scientists and engineers in order to slow down the develop-
ment of this new important scientific trend in our country.99

Thus, the coauthors held, the critics of cybernetics, not its proponents,
should be suspected of having fallen under the spell of the cold war enemy.
To recognize the contributions of the enemy without opening themselves
to attack, they heaped suspicion on suspicion, insinuating that instiga-
tors abroad had somehow organized the ideological critique of cybernetics
within the Soviet Union. Although it is unlikely that the coauthors genu-
inely believed that their discovery of cybernetics came in spite of the efforts
of American spies and agents, this kind of argument nonetheless won inter-
nal wars of words.
Soviet cyberneticists were not alone in employing this strained logic. If
Wiener was right in arguing that information arms all its possessors equally,
double heaps of suspicion may support an ultrarational strategy that strains
toward the irrationality found across cold war discourse. Kolman’s counter-
defense of cybernetics against other Soviet critics, for example, resembles
a game-theoretic scenario in which (like the policy of mutually assured
destruction) both parties seek to settle their disagreements in order to
avoid a larger collective loss.100 The basic logic of this cybernetic worldview,
asserts historian Peter Galison, is to adopt the logic of the “enemy Other”
and to preempt and predict the behavior of the intelligent and rational foe
to the point where the positions are reversed and foe and friend become
Cybernetics—like its sister disciplines of game theory, information
theory, and others—appears as a method for rationalizing the enemy, dis-
tributing structural strategy evenly across opponents and flattening the
chances that an enemy will have to take strategic or logical advantage

9800.indb 43 6/2/16 3:05 PM

44  Chapter 1

over an ally. Perhaps nowhere is this as clear as in the Soviet defense of

cybernetics itself, except that in Kolman’s case, the enemy to defend cyber-
netics against was his own kind. At first rejected for its American sources,
Soviet cybernetics took shape not as a Soviet reaction against the American
enemy but as a circular defense of Soviet mathematicians against their own

A “Complete Cybernetics”: Toward a Totalizing Plurality

The efforts of Sobolev, Lyapunov, Kitov, and Kolman in print and in public
lecture, combined with the intellectual weight of preeminent mathema-
tician Andrey Kolmogorov and high-ranking administrator and engineer
Aksel’ Berg, led to the establishment of the statewide Council for Cyber-
netics in 1959, which in turn promised cybernetics a base for significant
growth as an institutional field in the early 1960s. By 1965, however, it was
still not clear in which direction this new science would lead. Would it dis-
tribute the powers of the Soviet state among its participants more equitably
and flexibly? Or would it consolidate power still further? In 1965, an Ameri-
can visitor feared the worst: after visiting a facility with an evident gen-
eration gap between “all the young, recent graduates of technical higher
schools” who were interested in computers and “the older bureaucrats,” he
prophesied that “a turnover in generations in the Soviet administration”
could lead to a “computer revolution” that “may enormously increase the
effectiveness of formal communication channels.” The “modernization of
communication may have the paradoxical effects,” the American observer
fretted, “of actually enhance[ing] totalitarian control by making a fully
centralized network of administrative communications channels really
Between 1960 and 1961, the popular press began heralding comput-
ers as “machines of communism” and engineer admiral Aksel’ Berg, then
director of the Council of Cybernetics, launched the first of a series of vol-
umes entitled Cybernetics: In the Service of Communism.103 This series stirred
emotions among Western observers. One American reviewer noted with
concern in 1963, “If any country were to achieve a completely integrated
and controlled economy in which ‘cybernetic’ principles were applied to
achieve various goals, the Soviet Union would be ahead of the United States
in reaching such a state.” The reviewer also picked up on the burgeoning
interest in economic cybernetics, stating that “a significantly more efficient
and productive Soviet economy would pose a major threat to the economic
and political objectives of the Western World.… Cybernetics, in the broad
meaning given it in the Soviet Union,” he concluded with a flare, “may be

9800.indb 44 6/2/16 3:05 PM

A Global History of Cybernetics  45

one of the weapons Khrushchev had in mind when he threatened to ‘bury’

the West.”104
The Central Committee began publicly promoting cybernetics along
similar lines in 1961 at the Twenty-second Party Congress as “one of the
major tools of the creation of a communist society.”105 First Secretary Nikita
Khrushchev himself promoted a far-reaching application of cybernetics:
“it is imperative,” he declared to the Congress, “to organize wider appli-
cation of cybernetics, electronic computing, and control installations in
production, research work, drafting and designing, planning, accounting,
statistics, and management.”106 Central Intelligence Agency (CIA) sources
noted similar enthusiasm at an All-Union Conference on the Philosophi-
cal Problems of Cybernetics held in June 1962 in Moscow, which included
“approximately 1000 specialists, mathematicians, philosophers, physicists,
economists, psychologists, biologists, engineers, linguistics, physicians.”107
The conference adopted an official, if vague, definition of cybernetics as “the
science which deals with the purposeful control of complex dynamic sys-
tems.”108 The most ambitious of these complex dynamic systems, the Party
leadership’s support seemed to imply, would be the Soviet Union itself.
The looming menace of a well-organized, cybernetic self-governing
socialist enemy worried some American observers as well. During the John
F. Kennedy administration, members of the intelligence community agi-
tated against the perceived looming peril of Soviet cybernetics. John J. Ford,
then a Russian specialist in the CIA and a future president of the Ameri-
can Society for Cybernetics, was responsible for several alarm-generating
reports on Soviet cybernetics, which had already grabbed Attorney General
Robert F. Kennedy’s attention. One fateful evening in the fall of 1962, Ford
gathered with President John F. Kennedy’s top men to discuss the impend-
ing peril of Soviet cybernetics, only to have his meeting interrupted by
the announcement that surveillance satellites had just uncovered photos
of Soviet missiles in Cuba.109 By the time the dust settled after the Cuban
missile crisis, Soviet cybernetics no longer agitated the administration,
which had reviewed the science and did not deem it an urgent threat. It is
a strange twist of history, then, that the international crisis that is consid-
ered the zenith of cold war hostility (the Cuban missile crisis) also defused
and derailed mounting American anxieties about the “Soviet cybernetic
Although U.S. and Soviet intelligence officers alternately fretted about or
enthused over the possibilities of a cybernetically coordinated Soviet power,
the facts about the practical debates among Soviet scientists point in a very
different direction. Soviet cybernetics, for all its talk about self-governance,

9800.indb 45 6/2/16 3:05 PM

46  Chapter 1

was anything but. Berg’s series Cybernetics: In the Service of Communism pro-
duced heated debate and fierce divisions among prominent mathemati-
cians in the Soviet Union.111 In contrast to the CIA’s fear of a mounting,
unified platform of Soviet cybernetics, cybernetic talk swelled the internal
discord among mathematical cyberneticists, painting a picture instead of an
intellectually fractured front. Leading Soviet cyberneticists defined the field
in dramatically different terms: Kolmogorov fought to claim information
as the base of cybernetics, Markov preferred probabilistic causal networks,
Lyapunov set theory, and Iablonsky algebraic logic. In 1958, only three
years after their initial article, Kitov, Lyapunov, and Sobolev published an
article outlining four new definitions of cybernetics in the Soviet Union,
emphasizing the dominant study of “control systems,” Wiener’s interest
in “governance and control in machines, living organisms, and human
society,” Kolmogorov’s “processes of transmission, processing, and stor-
ing information,” and Lyapunov’s methods for manipulating the “struc-
ture of algorithms.”112 According to researchers, loose groups of cybernetic
thought consolidated around leading cyberneticists such as Lebedev, Berg,
Lyapunov, Glushkov, Ershov, and others.113
Although some scientists contended that the virtue of cybernetics lay in
its capacious tent of competing foundations, not everyone felt that the new
field should contain multitudes. Igor Poletaev, a leading Soviet informa-
tion theorist and author of the 1958 book Signal, an early work on Soviet
cybernetics, argued in 1964 against any plastic understanding of cybernet-
ics. He legitimated his call for disciplinary coherence by invoking its for-
eign founder, Norbert Wiener, claiming that “‘terminological inaccuracy’
is unacceptable, for it leads and (has already led) to a departure from Wie-
ner’s original vision of cybernetics toward an inappropriate and irrational
expansion of its subject.”114 “As a result,” Poletaev continued, “the specific-
ity of the cybernetic subject matter completely disappears, and cybernetics
turns into an ‘all-encompassing science of sciences,’ which is against its
true nature.”115 The geneticist Nikolai Timofeef-Ressovsky, whose life and
work was praised and persecuted under the regimes of both Hitler and Sta-
lin, once put the same sentiment in lighter terms. In correspondence with
Lyapunov, he replaced the Russian word for confusion or mess with the term
cybernetics, joking about his having once placed a letter in the wrong enve-
lope as a “complete cybernetics.”116 In Timofeef-Ressovsky’s witticism, we
uncover a fitting rejoinder to those enthused and worried that a complete
cybernetics might mean a unified Soviet information science and society.
To put it both precisely and audaciously, the term cybernetics should be
used in the plural, and perhaps the only stable sense of cybernetics is the

9800.indb 46 6/2/16 3:05 PM

A Global History of Cybernetics  47

adjectival form, cybernetic, an adjunct to anything that its users see fit to
apply it. From the point of view of the central committee that organized
cybernetics institutionally, Soviet cybernetics, at the peak of its reach,
appears both comprehensive and pluralist. It was a complete mess, as
Timofeef-Ressovsky jested. In the late 1960s, the Academy of Sciences of
the USSR promoted cybernetics into an entire division, one of four divi-
sions comprising all Soviet science.117 The remaining three (noncybernetic)
divisions—“the physico-technical and mathematical sciences, chemico-
technical and biological sciences, and social sciences”—could without
much conceptual violence be read as subfields of the Siberian-sized Soviet
cybernetic science.
The Soviets were not alone in the instinct to universalize science, although
the ideological organs of the state excelled at promoting such discourse.
The ecumenical commitment and a totalizing mission to stitch together
the mechanical, the organic, and the social often were attributed to their
foreign founder. In 1948, Wiener attempted to analogize (in the subtitle to
his 1948 Cybernetics) “the animal and the machine” and concluded with a
comment about the insufficiency of cybernetic methods for social sciences.
Nonetheless, two years later, in 1950, Wiener published a popular version
called The Human Use of Human Beings, whose subtitle belies his earlier cau-
tion: “cybernetics and society.”118 Still, the instinct to institutionalize his
intellectual catholicity was clearly native to the Academy of Science, which
originally categorized cybernetics into eight sections, including mathemat-
ics, engineering, economics, mathematical machines, biology, linguistics,
reliability theory, and a “special” military section.119 With Aksel’ Berg’s
sway over the Council on Cybernetics, the number of recognized subfields
then grew to envelop “geological cybernetics,” “agricultural cybernetics,”
“geographical cybernetics,” “theoretical cybernetics” (mathematics), “bio-
cybernetics” (sometimes “bionics” or biological sciences), and, the most
prominent of the Soviet cybernetic social sciences, “economic cybernetics”
(discussed in later chapters).120
By 1967, the range of cybernetic sections enveloped information theory,
information systems, bionics, chemistry, psychology, energy systems, trans-
portation, and justice, with semiotics joining the linguistic section and
medicine uniting with biology. Sheltering a huddling crowd of unorthodox
sciences, including “non-Pavlovian physiology (‘psychological cybernet-
ics’), structural linguistics (‘cybernetic linguistics’), and new approaches
in experiment planning (‘chemical cybernetics’) and legal studies (‘legal
cybernetics’),” cybernetics in the mid-1960s grew to an almost all-encom-
passing size.

9800.indb 47 6/2/16 3:05 PM

48  Chapter 1

Nonetheless, the runaway institutional success of cybernetics in the

Soviet Union also meant that, by the time Leonid Brezhnev came to power
in 1964, Soviet cybernetics could not help but slouch toward the intel-
lectual mainstream.121 It had to: its territory had grown so large it could
not help but take up the middle of the road. The institutional growth of
cybernetics outran the intellectual legs supporting it: the failure of cyber-
netics to cohere intellectually actually rested on the runaway growth of
the discipline institutionally. Sloughing reformist ambitions to the side,
by the 1970s, kibernetika signaled little more than a common interest in
computer modeling that held together a loose patchwork of institutions,
disciplines, fields, and topics. By the 1980s, the term cybernetics marked a
nearly empty signifier for all the plural things to which the adjective could
be attached. By the rise of Gorbachev in 1984, Soviet cybernetics had suc-
cessfully accompanied and slowly integrated into a host of parallel develop-
ments. The inheritor field “informatics,” the parallel revolution in military
affairs, the scientific-technical revolution, and the first three generations of
computer hardware (vacuum tubes, transistors, and integrated circuits) had
rolled forward under the fading banner of Soviet cybernetics.122

Conclusion: Wiener in Moscow

This brief history of early Soviet cybernetics ends where it began, with Nor-
bert Wiener and the foreign founding of cybernetics. In the early 1960s,
travel restrictions for Americans in the Soviet Union began to slacken, and
a trickle of chaperoned scientific and cultural exchanges began to flow
between the two superpowers. Early among this generation of guests was
Wiener, then an aging omnibus professor at MIT. In June 1960, Soviet offi-
cials warmly welcomed this American founder of cybernetics for a several-
week visit to Moscow, St. Petersburg, and Kiev (figure 1.1.) After his arrival,
Wiener, whose translated books were popular (albeit in edited form) in the
Soviet Union, was paid royalties in cheap caviar and champagne (which
apparently sat untouched in his basement) and gave invited lectures at
prestigious institutes in those three cities.123 For Wiener, it was a chance
to issue a stirring warning against societies that would adopt cybernetics
without the fundamental ability to correct themselves, decrying that “sci-
ence must be free from the narrow restraints of political ideology.” For his
Soviet hosts, the visit allowed the cybernetic knowledge base to go about
the regular ideological work of welcoming and canonizing a socialist saint
in public memory of Soviet society and technology.124 The effect among his
colleagues in the Soviet Union and in Cambridge was electric. Reflecting on

9800.indb 48 6/2/16 3:05 PM

A Global History of Cybernetics  49

his public reception, Wiener’s friend Dirk Struik, a Dutch mathematician

and Marxist theoretician, captured the moment for many Soviet cyberneti-
cists with his overstatement, “Wiener is the only man I know who con-
quered Russia, and single-handed at that.”125 We may claim that by this
process Wiener became known as a foreign founder of Soviet cybernetics.
In Democracy and the Foreigner, political theorist Bonnie Honig introduces
the idea that an iconic “foreign founder,” or an alien recruited for a project
that he or she unsettled, often plays a role in the many political narratives
of identity formation: the kingdom of Oz has its Dorothy of Kansas; the
House of David has a Moabite grandmother, Ruth; the American colonies
were united by the belief that they were no longer British; Europe now
traces its origins to ancient Greece, which was first a Roman idea. Eastern
Europe abounds in similar stories: Russia originates in ancient Rus’, now
in Ukraine; the Ukrainian national anthem claims brotherhood with the
Cossack; and the Polish national anthem praises Lithuania.126 That Soviet
cybernetics identified Wiener as foreign founder is in context nothing
new. After all, no native can found his or her identity. There is no identity

Figure 1.1
Norbert Wiener with Aleksei A. Lyapunov in Moscow, 1960.
Courtesy of Boris Malinovsky.

9800.indb 49 6/2/16 3:05 PM

50  Chapter 1

without a founder, and because founders precede identities, all foundations

must be laid by what must appear post fact as foreigners.
Wiener’s renown in the former Soviet territories has outlasted his mem-
ory in the English-speaking world. When Aksel’ Berg became chair of the
Council on Cybernetics in 1959, he made sure that among the first support-
ing works translated were Wiener’s. Over fifty years later, nearly all of Wie-
ner’s major works have since been translated into Russian and retain their
relative popularity, long after his legacy has faded in the English-speaking
world, except recently among historians of science.127 Wiener’s 1948 Cyber-
netics, or Control and Communication in the Animal and the Machine was trans-
lated into Russian in 1958 and reissued in 1968 and again in 1983, one more
printing than in English. In 1958, his Human Use of Human Beings: Cyber-
netics and Society was abridged and translated as Kibernetika i obscheshtvo
(Cybernetics and Society). Based on the lectures he gave while visiting Mos-
cow, he published a 1962 article “Science and Society” in the preeminent
journal Problemy Philosophii (Problems of Philosophy). His autobiographies
Ex-Prodigy (1953) and I Am a Mathematician (1956) were translated in 1967.
And his final collection of lectures, God and Golem, Inc.: A Comment on Cer-
tain Points in Which Cybernetics Impinges on Religion (1964), was translated as
Tvorets i robot (Creator and Robot) in 1966 and reissued in 2003.
As a testament to the staying power of Wiener as an iconic foreign founder
figure, Wiener’s semiautobiographical novel The Tempter was translated in
1972, eight years after his death. His short piece of fiction, “The Brain,”
which is hard to find in English, was translated in 1988. And his 1951 article
“Homeostasis in the Individual and Society” appeared in Russian in 1992,
just after the turbulent collapse of Soviet society. Bookstores in Moscow con-
tinue to offer new editions of Wiener’s works to this day. His oeuvre has also
migrated online unevenly: all aforementioned works in Russian are freely
available for download online, compared to only one work in English, God
and Golem, Inc. Given all this, it may not be a stretch to assert that, with the
visit of an American founder of cybernetics, the son of Leo Wiener, an émigré
from Byelostock and founder of Slavic studies in America, Norbert Wiener
was christened no less than a Soviet prophet returning home.128
Yet if Wiener were a prophet, he would be the kind whose stinging
calls to repentance went ignored both at home and abroad. He pressed for
removing ideology from science just as the political winds, in the early
1960s, were shifting toward ideological reconsolidation and recentraliza-
tion under Brezhnev. The case of Wiener in Moscow is interesting, then,
not merely for biographical or historiographical reasons but also as a syn-
ecdoche for the larger Soviet experience with cybernetics. The cybernetic

9800.indb 50 6/2/16 3:05 PM

A Global History of Cybernetics  51

technological apparatus brought with it a promise of systemwide structural

reform, and although that reform was never fully realized, the technologi-
cal apparatus was. Cybernetics accompanied the transformation of Soviet
society into an already networked information society, although it did so
without bringing about the intended social, organization, and techno-
logical reforms and self-governance. The early nationwide cybernetworks
explored in subsequent chapters are central to understanding the Soviet
experience and the unintended political consequences of sociotechnical
and technocratic reforms.
A glance at the history of early Soviet cybernetics might at first steer
readers to think that technocratic sciences are politically neutral, capable
of adapting to whatever the political discourse of the day is, whether Sta-
lin’s rejection, Khrushchev’s reform, or Brezhnev’s reconsolidation of tech-
nocratic science. Yet this is not the case: claiming technocratic neutrality
itself is a consequential political posture that often is filled by whatever the
politics of status quo at the time and place are. The nationwide networks
created to save the flagging economy and technical data infrastructures dis-
cussed in later chapters are presented as socially neutral technocratic solu-
tions to social problems—and yet that position of neutrality proved to be a
veiled form of ideational investment. Considered generally, the cybernetic
goal of controlling and regulating information systems in abstract and sup-
posedly neutral mathematical terms appealed to post-Stalinist scientists
who were fed up with political oppression. Cybernetics struck Moscow-
based bureaucrats and party officials as a politically feasible way forward
in preserving the centralized state as an information system without the
abuses of Stalinism.129 Behold the promise of control without violence and
of a socialist information society liberated from its stained past by the neu-
tralizing politics of computation.
Others promised technological improvements without politics long
before the onset of computers and digital media. Soviet discourse of what
James Carey called the “electric sublime” begins with Lenin’s famous 1920
statement that “Communism is Soviet power plus the electrification of
the whole country,” perhaps the highpoint of the Soviet reputation in the
West as well as a memorable declaration of the Soviet Union’s commitment
to achieve social progress through technological modernization.130 Soviet
cybernetic discourse built actively on that tradition—particularly that of
the Soviet digital economic network projects, which, like Lenin’s electrifi-
cation (or GOERLO) project, promised to rework the technological infra-
structure of the whole country—the factories, the grids that united them,
and the giant hydroelectric and computer stations that powered them. The

9800.indb 51 6/2/16 3:05 PM

52  Chapter 1

cybernetwork projects integrated and updated a longer tradition of the

industrialist, Taylorist megaprojects that marked the Soviet electrical age.
The cybernetic lexicon also resonates richly with native Soviet discourse.
Before Wiener cemented that hardy word as central to cybernetic systems,
feedback occupied a prominent position in the Soviet political imagination
of itself as a “socialist democracy,” a kind of complex social entity sus-
tained by Pavlovian mechanisms of stimulus and response and control and
cooperation between rulers and masses.131 With little work, the term noise
reduction came to stand for a technical synonym for continuing political
censorship in the Soviet Union. Moreover, Wiener’s twinning of the mod-
ern laborer with an automaton echoed of Stalin’s attempts to make Soviet
labor and industry efficient with the scientific management techniques of
Taylorism. Wiener’s theories of systematic information control and com-
munication, once translated into Russian, appeared to be a recuperation of
ideas that already were well understood.132
Perhaps this history of Soviet cybernetics is most helpful not for what
it says about cybernetics but for what the discursive pliancy of cybernetics
allows us to see in Soviet society. As a term, cybernetics served as a flexible
semantic placeholder for a more widely held article of faith about the prom-
ise of technocratic governance aided by computer in post-Stalinist science
and society. As a history, the several-step process of the Soviet rejection,
rehabilitation, adoption, and adaptation of a new foreign discipline reveals
less about cybernetics than it recapitulates the preexisting political dynam-
ics of Soviet discourse—the debate patterns, rituals of discourse, strategies
for intellectual defense, alliance forging, institution building, the political
whims of Moscow, and other everyday dynamics. Backlit with fascinating
twists, turns, and figures, the story of Soviet cybernetics presented here
signals not particularly well-defined intellectual contributions but rather
shows the ways that the lack of them allowed Soviet cybernetic discourse to
mold to and reflect longer transformations and trends in the Soviet state’s
attempts to manage and control science, technology, and society.
Soviet cybernetics thus appears to be a normal science in the sense that
it reveals the conflicting dynamics of underlying political, economic, and
institutional practices and structures. These dynamics—the echoes of anti-
capitalistic public campaigns, the ritual aspects of intellectual debates and
duels, the political machinations and strategies, the institutional diffusion
of the computer as a specialized tool, the history of spikes of invention
followed by downward-sloping plateaus of innovation and development
characterizing the history of science in Russia, and the stubborn fact that
the work of science takes place in prolific dialects and varied trading zones

9800.indb 52 6/2/16 3:05 PM

A Global History of Cybernetics  53

subject to the punishing pleasures of contest, prestige, and competition133—

appeared par for the course. As the next chapter attempts to illustrate, the
case of Soviet economic cybernetics challenges historians and other agent-
observers of change with the suggestion that perhaps the ordinary, over-
looked elements of actors, ideas, practices, and policies—including those
governing everyday life in the command economy—best describe the cir-
cuitous historical course of science and social reform.
In one important sector, however, Soviet cybernetics and other infor-
mation sciences were not obviously subjected to a confusion of com-
peting motives—the Soviet military. The Red Army adopted cybernetic
research methods and vocabulary, usually coded in public simply as “spe-
cial research”; successfully theorized the military-technical revolution
spurred by computers and associated long-range, specific-target military
innovations; and maintained a competitive space and nuclear and long-
range conventional warfare armaments without the internal incoherence
and competition that was found in civilian sectors. So although the Soviet
cybernetic-lit military technology revolution of the 1970s did not lead to
application due to the political and economic incapacities of the Soviet
state, the key distinction from the civilian economic sectors is that, inside
the centralized military administration, real cybernetic reform was both
possible and carried out in theory.134
In conclusion, having outlined a few sources that led to the consolida-
tion of cybernetics in Wiener’s 1948 masterwork, the Macy Conferences
on Cybernetics (1946–1953), its postwar spread through France, England,
Chile, and a vignette of how cybernetics became a loose techno-ideologi-
cal framework for thinking through information sciences in post-Stalinist
Soviet Union, I now comment on the idiosyncratic development of cyber-
netics across these moments in the early cold war global history. Several
comparisons and contrasts draw connections to other postwar climates
where cybernetics came to roost. The Soviet translation and adoption of
cybernetics share with the other case studies glossed here an underlying
fascination with the relationship of the mind to the machine, especially
as seen in the biology and neurology of the British and Chilean cybernet-
icists. The mind-machine analog is a politically charged two-way street.
Not only does cybernetics prompt us to think about how a logic machine
(computer circuits or any other Turing machine) may function like a mind
(a neural network), but it also raises McCulloch’s potent possibility that
subsequent neuroscience has soundly rejected: the mind (neural network)
might function like a logic machine (computer circuits). This reverse com-
parison (that a mind is like a machine) proves particularly enduring in

9800.indb 53 6/2/16 3:05 PM

54  Chapter 1

later discussions of the design and development of national networks. The

designers of major early cold war national networks in the Soviet Union,
the United States, and Chile sought, implicitly or explicitly, to model their
own self-governing national networks after cybernetic neural networks. In
the comparative network designs (including distributed, hierarchical, and
participatory), early network scientists proposed differing images of the
relationship between a network and the living body politic of the nation.
The mind analogies all share a common cybernetic impulse to analogize
between information systems underlying organisms, machines, and societ-
ies. (The organizing itch of cybernetics is simply that a better-understood
system can inform a less well-understood system.) Analogies are neither
right nor wrong: they should be judged by their interpretive use rather
than their epistemic weight. (Or as Evelyn Fox Keller once noted, the word
simulation meant deception before it meant analogical likeness.) Given this,
it is striking that each of the network architect teams at hand (Glushkov’s
OGAS, Beer’s Cybersyn, and Baran’s ARPANET) chose to analogize or model
its national network project after the same basic image—the human mind,
or an organic nervous system. But each of these national networks expressed
the basic design analogy differently.
These early national networks projects—OGAS, Cybersyn, and ARPA-
NET—were designed after different models of the (human) mind. Even
though Beer and the Cybersyn project rejected previous and ongoing Soviet
attempts to manage the command economy, the OGAS and Cybersyn proj-
ects pursued a national model in which the nation is likened to the body of
an organism and the computer network to the nervous system that incorpo-
rates that nation’s communication.134 The ARPANET, by contrast, inspired
by McCulloch’s neural network research, is analogized to a disembodied
brain itself. In this case, the nation is like the brain itself: whatever organi-
zation the network serves constitutes its own neural network. To oversim-
plify, Baran foresaw a national state network simulating a brain without a
body, while Glushkov (and Beer) anticipated a network nation simulating a
body with a brain—a government in touch with its people.
As cognitive philosophers have submitted, analogies of (a nation as) an
embodied mind and a disembodied brain work very differently. Although
Soviet scientists were understandably wary of overbold political proclama-
tions, the OGAS design reaffirmed the self-conception of the Soviet state
as a decentralized hierarchical heart of the Soviet nation. In a colossal
nation, workers were to be incorporated and animated by planning that
emanated from the central processing unit, or social brain, in Moscow. That
state would not be simply centralized and top-down. In the OGAS design,

9800.indb 54 6/2/16 3:05 PM

A Global History of Cybernetics  55

the network would serve as a nationwide nervous system that responded

to and adjusted in real time to local events and maintained dynamic bal-
ance through complex feedback loops with its internal and international
information environment. This metaphor was both materialist and ideal-
ist—materialist in that it grounded the nation in the industrial and eco-
nomic realities already on the ground and idealist in that it ignored the
fact that the economy did not behave like a healthy or single body (but
instead like an environment for nonsymbiotic competition over bureau-
cratic positioning).
Also consider how the U.S. ARPANET analogy of the nation as a disem-
bodied brain, although articulated here for the first time to my knowledge,
has already been inscribed many times. Most often the interpretations
smack of triumphalist political overtones. Seeing the nation (network) as a
brain, not a body, signifies that the United States is conceived as an organ
for knowledge work, not physical labor; that its civilian communication
networks imagine its citizens, not the state, as the democratic decision-
making mechanism for the nation; that those citizens exist in peer-to-peer
relationships where, like nodes in a distributed network, each may act and
interact with her neighbor as equals; and that (particularly common in digi-
tal libertarian discourse) the computer network itself constitutes the higher
order of technological freedom that is necessary for the natural emergence
of a more robust political order. (When Baran described distributed net-
working, his word was not robust but survivable because his network was to
survive nuclear attack, which puts a less optimistic spin on things.) Baran,
we might assert, was acting in the libertarian tradition by espousing the
organic nation as a marketplace of individuals dating back to Herbert Spen-
cer.136 Or perhaps Baran designed the ARPANET after the image of the state
as an enlightened social brain, channeling the American progressive notion
of the state (or any other depository of organized intelligence, including
the news-reading public, schools, universities, and scientific laboratories)
as a “social sensorium” dating back to John Dewey and Charles Horton
With enough imagination, the analogy of the national network to a
human mind can serve almost any end, such as the engine of a sensing
being interacting in a mediated environment, a nervous system animat-
ing a living body, or the gray matter filling a skull. Perhaps the reason that
these cyberneticists populated their analogies with the human mind was
(to paraphrase a leading neuroscientist) simply the fact that humans like
to believe that the human mind is the most complicated thing in the uni-
verse, even though this idea is probably no more than the brain’s opinion

9800.indb 55 6/2/16 3:05 PM

56  Chapter 1

about itself.138 The point is that this analog, like all others, is contentless.
It has no right or wrong, and the work that it does for us is work that
we do to ourselves. The stories we tell ourselves about our networks reveal
more about us—the spinners of modern-day network rhetoric—than it does
about the network itself.
Finally, this chapter summarizes how, in its early adoption period, early
Soviet cybernetics muted but did not erase politically potent mind-nation-
network questions with language that was deliberately more technocratic
and theoretical perhaps than that of cyberneticists in other countries.
Although no surprise, talk about cybernetics and society took on the tech-
nical discourse of what Gerovitch calls Soviet “cyberspeak,” or an ideologi-
cal and discursive strategy for embedding public discussion about society
in the language of technical expertise. The postwar and cold war debates
about cybernetics in the Soviet Union impinged on the social implications
of the new science. Perhaps the most obvious example of a technocratic
approach bearing out social implications is the focus of the next chapter—
the case of economic cybernetics. How, if at all, might cybernetics—or the
study of communication systems that organize our bodies, machines, and
societies—improve the current social, political, and economic order? How,
as Stafford Beer developed in Chile, might cybernetic insights be applied to
the networking of the Soviet nation in need of an economic boost? How
might concerns with communication and control that were central to both
the larger Soviet state and cybernetic projects play out in the crucial prac-
tice and policies of command economies? The following chapter discusses
these and other questions.

9800.indb 56 6/2/16 3:05 PM

2  Economic Cybernetics and Its Limits


Civilian computer networking in the Soviet Union first developed among

cyberneticists who applied their science to a unique environment—the
command economy. By examining the work of economic cyberneticists—a
field found only in the territories of the former Soviet Union—we can begin Econom
to understand the significance of the internal economic crisis to Soviet sci-
entists and civilians and the ways in which Soviet scientists, administrators,
and policymakers in 1959 to 1963 viewed the command economy itself as
a complex cybernetic organization. In this light, the same terms were used
both by key Soviet network entrepreneurs to envision the first national
networks as well as by the critics who condemned those projects. By review-
ing the organizational theories and practices that characterize the Soviet
state socialist economies, this analysis explores and begins to complicate
the divide between the private markets and the public states that underlie
conventional conceptions of the cold war.1
The command economy contained in its operations the cybernetic seeds
and complex sources of its own undoing—nonlinear command and con-
trol, informal competition, vertical bargaining, and what I am calling heter-
archical networks of administrative conflict. In this chapter, I develop these
observations through a series of examples that outline the basic operations
of the command economy in theory and in practice, the various schools
of thought concerning economic reform (especially around the transition
from Nikita Khrushchev to Leonid Brezhnev in 1963), and the political
tensions that economic cybernetics tried to square itself with in an attempt
to reform (often with long-distance networks) the structural contradictions
underlying the practices of the command economy. These contradictions
slowed efforts at technocratic economic reform and also ensured the endur-
ing appeal of nonlinear cybernetic systems thinking.
The term command economy originated from the German Befehlswirtschaft,
which was used to describe the Nazis’ centralized economy and socialist

9800.indb 57 6/2/16 3:05 PM

58  Chapter 2

economy. A command economy is one in which the coordination of econ-

omy activity is carried out not by market mechanisms but by administrative
means through commands, directives, targets, quotas, regulations, and the
like.2 Karl Marx and Friedrich Engels said almost nothing about economic
planning, except that it would be necessary, and Engels left the decisions to
the workers.3 They also asserted that socialism would be impossible to build
in impoverished societies, which Leon Trotsky associated with tsarist Russia
before fleeing to Mexico. Nikolai Bukharin foreshadowed what followed
next when he said that “as soon as we make an organized social economy,
all the basic ‘problems’ of political economy disappear: problems of value,
price, profit, and the like. Here ‘relations between people’ are not expressed
in ‘relations between things,’ and the social economy is regulated not by
the blind forces of the market and competition, but consciously by a …
plan.”4 On such promises, the Russian revolution was built. Nonetheless,
from 1917 until the collapse of the Soviet state in 1991, a perennial puzzle
dogged the Marxist-Leninist state planners: How precisely was that plan
supposed to work? How can a state command an economy?
Some tenets of the Soviet answer are clear.5 All the means of industrial
production were nationalized, and although Soviet citizens could own
some “individual” (not “private”) property (including houses, apartments,
and automobiles), few could afford to do so.6 The Soviet state appointed
three state ministries to serve as the nation’s economic brains, budget-
keeper, and managers of the nation’s vast property holdings and means of
production—the Gosplan (State Planning Commission), the Gosbank (State
Bank), and the Gossnab (State Commission for Materials and Equipment
Supply). (Gos is short for gosudarstvo or Russian for state or government.)
Gosbank, the central bank that prepared the state budget with the Ministry
of Finance, played a transactional accounting role and the least critical role
of the three.
Gosplan and Gossnab carried out crucial and different roles. Gosplan
was entrusted with creating the economic plans of action—the governing
documents defining the economic inputs (such as labor and raw materials),
the timetable for execution, the wholesale prices, and most of the retail
prices—divided into five-year increments (the so-called five-year plans).
These nationwide economic plans were first rolled out from 1929 to 1933
under Stalin and ended, with one seven-year exception (1959–1965) under
Khrushchev, with the twelfth plan (1986–1990), which oversaw Mikhail
Gorbachev’s reform policies of uskorenie (acceleration) and perestroika
(rebuilding). The thirteenth five-year plan was cut short by the dissolution
of the Soviet Union in 1991.

9800.indb 58 6/2/16 3:05 PM

Economic Cybernetics and Its Limits  59

Gossnab, in contrast, was responsible for implementing Gosplan’s plans

by procuring and supplying producer goods to factories and enterprises
and by monitoring the schedules for the production plans. Gossnab thus
fulfilled the market role of allocating goods to producers and bridged the
three levels of the command economy—national, regional, and local plan-
ning and production. The three-tiered model, established under Stalin in
the 1930s, presents a straightforward pyramid. Gosplan sat at the top level
and politically determined the national targets for each sector and indus-
try, those targets are divided hierarchically among the midlevel of regional
ministries, and they are further subdivided at the bottom level among
enterprises and factories themselves.7 If Gosplan planned it, Gossnab car-
ried it out across all three levels—or at least that was the plan. As I lay out
below, Soviet bureaucrats came to understand that at its heart, Soviet eco-
nomic planning was a cybernetic process. This understanding goes a long
way toward explaining the curious fact that the same state planners and
economic agents later resisted attempts to implement large-scale cybercom-
puting networks in the Soviet Union.
The Soviet command economy grew at tremendous human, environ-
mental, and organizational costs. In wartime, the command economy
worked well enough to survive the extreme national duress of World War II,
in which a devastating 26 million people or 14 percent of the Soviet popu-
lation perished between 1941 and 1945. For the next few decades, Soviet
gross national product grew faster than elsewhere in the world, enjoying
a peak growth rate of 7 percent in the 1950s and 4 to 5 percent in the
1960s (before flattening out to a 2 percent growth rate in the 1970s and
finally stalling at zero in the 1980s). In 1987, the “oppositionalist” Soviet
economist G. I. Khanin estimated that Soviet economic productivity grew a
total of 6.6 times (not the official claims of 84.4 times) since 1928—which
by raw indices alone, is a history of economic growth similar to normal
industrialized economies.8 By far the most unforgivable and unforgettable
cost to Stalin’s rapid pace of economic development came in human lives.
Some estimate that as many as 10 million lives were lost, many of them
forced famine victims, surely among the most despairing statistics in mod-
ern history.9
Stalin built the state at inhuman cost, but he built it nonetheless. Under
Lenin’s and Stalin’s leadership, the command economy modernized a pre-
industrial country that was run by a few into a mighty industrial power.
It began in 1917 with a small group of professional socialist revolutionar-
ies who lived in a few cities in a huge country that was 84 percent rural
and whose population was over 95 percent illiterate peasants. After their

9800.indb 59 6/2/16 3:05 PM

60  Chapter 2

October coup, the Bolsheviks eliminated the remnants of the oppositional

armies run by the tsar and the Mensheviks, among others, and developed
an advanced industrialized economy that, after a couple of decades of
forced modernization, helped the Allies defeat the Nazi war machine. As
the cold war ensued, the Soviets, fueled in part by state paranoia and in part
by scientific ambitions, maintained military parity with the United States,
obtaining nuclear energy and weaponry before most of the rest of the world
and pulling ahead in the space race in the late 1950s.
The political economy was also engineered to advance meaningful
civilian causes such as socioeconomic justice. In most empires, the reve-
nue flows from colony to center, but in the Soviet Union the funds ran in
reverse: Moscow invested more in supporting satellite republics and regions
than it stripped from them. The state mandated education, raised literacy
rates for millions, granted women skilled and technical positions in the
workplace, and successfully exported huge amounts of natural resources,
ensuring a Soviet presence on the international economic stage. In the
1920s, before the Great Depression and before the 1930s purges, the gulags,
and other Stalinist abuses became widely known, most intellectuals in the
West admired at least some parts of the ambitious social projects that rode
the coattails of the Russian revolution.10 Optimism glimmered again after
the death of Stalin in 1953 and through the heady years of the early 1960s,
when all outside indicators suggested that the magic of the command econ-
omy—a fairytale on which a repressive empire had been built—might actu-
ally be working.
Yet those backstage had a better view of the problems. The degree of
information coordination between Gosplan and Gossnab—the brain for
planning and the hands of the command economy—was taxing the peace-
time state administration. Many things could go wrong and did. Gosplan
planned it, but Gossnab did not follow through. Or Gosplan planned
wrongly so that, even when properly executed, the plan did not meet
the economy’s needs. Rarely, if ever, did the command economy work as
The problems that economic planners and practitioners faced multiplied
in application. They include an accounting burden accumulated from inno-
cent calculation errors, compounded incentives that distorted reporting,
toilsome paperwork, structural inconsistencies across industry standards,
prohibitively technical product orders, uncoordinated silos of the national
planning apparatus already awash in pricing decisions and administrative
deluge, and many other practical problems that manifested themselves to

9800.indb 60 6/2/16 3:05 PM

Economic Cybernetics and Its Limits  61

cyberneticists and other economic planners as informal competition in the

command economy.
The institutional map of the command economy grew labyrinthine as
the immense accounting burden—a hulking coordination problem (or in
cybernetic lingo, an information-processing problem)—that was shoul-
dered by Gosplan and Gossnab was complicated by the participation of the
Ministry of Finance, the Central Statistical Administration, and the Minis-
try of Defense (defense is thought to have occupied as much as one quarter
of the USSR’s GDP in the late 1980s, although estimates vary widely).11
In the first six months of 1962, the priority industries that produced steel
tubes, mineral fertilizers, agricultural machinery, chemicals, oil, cement,
and light steel fell to at least 7 percent under quota—which some criti-
cal accountants attributed to human calculation errors. A calculation error
could mean too low production targets for heavy machinery one year, and
too little heavy machinery that year meant cross-industry shortfalls the
next.12 Even growth, when unforeseen, spelled trouble: in 1962, it was dis-
covered that the ongoing seven-year plan had overlooked the 1959 census
data and that by July 1962, the Soviet population had grown by 4 million
more than had been planned for. Khrushchev once predicted that the pop-
ulation discrepancy by the late 1960s would border on 15 million people
unaccounted for in the young, nonproductive workforce.
Even the best-laid plans, no matter how accurately made at the ministry
level, went awry from ministry to regional council to factory. On the fac-
tory floor, people encountered widespread problems when translating the
quotas and orders into day-to-day operations. One factory was known for
decades after the war as the producer of a series of increasingly obsolete
automobile models, including the luxury government limousine known
as the ZIL (the abbreviation for Zavod imeni Likhacheva). The ZIL factory
(or Likhachev factory) received orders and quotas that were so specialized
that they required especially trained experts to interpret and execute. Yet as
investigators discovered in the early 1960s, only two out of sixty-four fac-
tory employees had any higher education, and twenty out of sixty-four had
not completed high school. Few on the factory floor could read, yet alone
fulfill, the specialized orders they received.
Every information-planning problem was also a coordination and thus
organizational-institutional problem, and the further up the economic
hierarchy, the more intractable the coordination problems. Even at the top
of the ministries, the economic plan did not necessarily exist in a single
coordinated document, and so silos of attention regimented and splintered

9800.indb 61 6/2/16 3:05 PM

62  Chapter 2

the planning process. Consider this 1962 complaint in Pravda from a fac-
tory director about the determination of cost:

The department of Gosplan that drafts the production program for Sovnarkhozy
[collective farms] and enterprises is totally uninterested in costs and profits.… Ask
the official in the production programmed department in what factory it is cheaper
to produce this or that product. He has no idea.… He is responsible only for allocat-
ing production tasks. Another department, uninterested in costs, decides the plan
for gross output. A third department or subdepartment proceeding on the principle
that costs must always decline and productivity increase, plans the costs, the wage
fund, and the labor force on the basis of past performance. Allocations of materials
and components are planned by numerous other departments. Not a single depart-
ment of Gosplan is responsible for the consistency of these plans.13

Some ministries tried to address these problems by tailoring their own

plans in-house. For example, the Ministry of Wood and Wood Process-
ing streamlined and unified the procedural notation for its medium-sized
industry. The resulting code, once formulated and printed, weighed in at
a wrist-breaking eighteen hundred pages and proved incompatible with
other industries.14
Given such perpetual misfits between plan and practice, the Soviet
search for the “perfect” economic organization was, in Gertrude Schroeder’s
understatement, “continuous.” The annals of Soviet economic planning
match decade after decade of bold conceptual innovations with perpetual
practical setbacks. The Gossnab ministry itself was dissolved or recreated at
least once every decade after its creation in 1947. It was fully dissolved in
1953 after Stalin’s death; was recreated in 1965 under Brezhnev, where it
oversaw the delivery of over two thousand essential products; underwent
various shufflings of responsibilities; and finally was stripped of the politi-
cal supply of petroleum products in 1981.15
All in all, the coordination problem was simple to state yet bewilder-
ing to solve: how could the nation best manage, harmonize, and organize
all the information variables, planned and otherwise, that were flowing
through its economy? How, if at all, could the Soviet knowledge base—
including economic cyberneticists, a group known for a taste for circular
problems—hope to account for the deficiencies of accounting in the sys-
tem? In 1962, the State Committee for Automation and the Institute of
Statistics estimated that roughly 3 million citizens (about 1.3 percent of the
220 million total) were engaged in public accountancy, data registration,
statistical and planning calculations, and other supporting information ser-
vices for the planned economy and that the number was rising fast. And
yet no one, outside of strong-armed national commanders under extreme

9800.indb 62 6/2/16 3:05 PM

Economic Cybernetics and Its Limits  63

wartime conditions, could manage and execute all the operations neces-
sary to sustain the administrative creep of bureaucrats that was necessary to
oversee the businesses, factories, and industries that were driving a national
economy. In 1962, Viktor Glushkov, the prominent cyberneticist and archi-
tect of the OGAS Project, formulated the problem that his network project
proposed a cybernetic solution for: he estimated that if the current paper-
driven methods continued unchanged, the planning bureaucracy would
grow by almost fortyfold by 1980, requiring the entire adult population of
the Soviet Union to be employed in managing its own bureaucracy.16

The Many Pathways and Pressures to Reform

Under Stalin’s centralizing rule, the pressures to reform the cumbersome

bureaucracy of the command economy were immense yet bottled up. At
21:50 on March 5, 1953, Iosif Vissarionovich Dzhugashvili, or Stalin (a
portmantetau of Russian stal or “steel” and “Lenin”), died of an apparent
brain hemorrhage. His was possibly the most consequential death of the
twentieth century. It set off waves of economic reform. A mere ten days
after he died, as a salute to their deceased strong leader and out of a gut
instinct for damage control, the Politburo combined twenty-four minis-
tries into eleven strengthened ones. The reformer Nikita Khrushchev would
have been foiled from implementing systematic administrative and eco-
nomic reforms because the reforms had begun before he could ascend to
power as the new general (and then first) secretary: under his administra-
tion, a series of uneven and troubled reforms were enacted between 1956
and 1965.
By the time that Khrushchev secured power, the winds of administrative
reform were blowing in the opposite direction (even administrations fol-
low dialectical patterns). Beginning in 1954, he began introducing dramatic
reforms to decentralize Stalin-era control over the economy, ceding some
Kremlin power to national, regional, and local subcommittees. Gossnab
itself—the national ministry for allocating goods—was dissolved from 1954
to 1964. In 1955, new laws significantly broadened the powers of regional
and local planning councils, leaving in their hands for the first time in
decades questions about their own financing, planning, capital investment,
labor and worker pay, and even some cultural and social projects. Factory
directors also took more direct responsibility in determining their factory’s
planning, financing, and pay situation. In 1957, Khrushchev did away with
national industrial ministries and replaced them with regional economic
councils (called Sovnarkhozy). He continued to implement similar measures

9800.indb 63 6/2/16 3:05 PM

64  Chapter 2

over his years in power, further splintering and territorializing the single
national economic administrative hierarchy into 105 economic-admin-
istrative regional councils that were overseen by ten general and fifteen
union-republic ministries. The 1957 economic decentralization, Khrush-
chev hoped, would help streamline and localize the planning process for
a monstrously complex and administratively top-heavy postwar economy
with over 200,000 industrial enterprises.
The causes and effects are hard to sort out. It is estimated that of the 44.8
million workers in the Soviet Union in 1954, the administrative personnel
made up 6.5 million of them, or 15 percent of the national workforce.17
No doubt Khrushchev also harbored some hopes that his decentralizing
reforms would release him from bearing sole responsibility for the health
of the whole Soviet economy. And yet the reforms did not work as hoped:
GNP growth plunged from 8.4 percent in 1956 to 3.8 percent in 1957, the
year of Khrushchev’s major reforms, and bounced around a 5 percent aver-
age until the Khrushchev-toppling disaster that was the poor harvest of
1963 (-1.1 percent decline, the only year with a negative GDP growth until
the end of the Soviet Union).18
Cybernetic economists quickly learned a point that network theorist
Alex Galloway has subsequently clarified: control does not necessarily dis-
sipate with decentralized or distributed networks.19 It exists in the protocols
and the (network) administrators and their rulings, and planning protocols
were periodically scrambled. Instead of accounting production by volume,
piecemeal targets were set after decentralized planning decisions. Instead
of empowering and streamlining the local economy, the decentralizing
reforms enraged the old guard in Moscow against its reformer and enlarged
the nation’s economic administrative apparatus. The overwhelming politi-
cal effects of widespread decentralization among economic administrations
alienated and frustrated many party officials, exacerbating the disarray and
discontent already attached to Khrushchev’s volatile leadership. Nonethe-
less, Khrushchev’s decentralization allowed for several schools of economic
thought in the early 1960s to percolate into public discussion and to cohere
in the debates among the top party leadership about the best path of reform.

Orthodox, Liberal, and Cybernetic Economists

In the transition from Khrushchev to Brezhnev, several camps (schools

presumes too much order) of thought coalesced around the question of
economic reform. The first camp included a generation of orthodox econ-
omists who clung to positions that many had gained under Stalin, held

9800.indb 64 6/2/16 3:05 PM

Economic Cybernetics and Its Limits  65

the then contemporary functioning of the command economy (with its

pyramid bureaucracy of paper stretching across national and regional plan-
ners, accountants, and quotas), and was not only doing just fine but was
the only ideologically approvable means for advancing socialist economy
toward communism. The most severe of the antirevisionists had long put
forward a rearguard defense of their own positions in power, which was
the historical paradox that any reform to the political system would be an
unacceptable deviation from the original Marxist project. Not even Marx
knew how such an economy would work, and his Soviet legacy was one of
continuous political economic reform. Even ultraorthodox economists had
trouble persuading others that there was no room for any economic reform
in the wake of Khrushchev’s own economic reforms, the political thaw, and
unstable economic growth. With so much at stake, no one could disagree:
something had to change. The orthodox economists had to concede that
there was room for debate.
The second camp took up what later was called the liberal economic posi-
tion and came onto the public scene in early September 1962 with the pub-
lication of a Pravda article by a once obscure economics professor, Evsei G.
Liberman. Liberman, the youngest son of a Ukrainian Jewish forest guard
from Galicia (who eventually emigrated to New York), came to the field of
economic planning relatively late at the age of thirty-seven while visiting fac-
tories in Germany in 1933. He also was responsible for introducing punched-
card computers—Powers and Hollerith perforating machines—for planning
in Ukrainian factories. In that 1962 Pravda article (titled “Plans, Profits, and
Bonuses”), Liberman introduced the signature piece of his reform platform,
the idea of profit reform, which he had developed in his 1956 dissertation
“Profitability of Socialist Enterprise.” Liberman offered up a galvanizing call
for economic reform—one that would require little more than the stroke
of a pencil and a slight retooling of the planning apparatus.20 He proposed
that the efficiency of an economic enterprise should be measured by its
profitability rather than its output, that profit measures would encourage
production efficiency and quality, and that profitable enterprises should be
incentivized by increased salary and bonuses. Liberman’s proposal initially
gained the support of Vasily Nemchinov, a leading Soviet economist-mathe-
matician and early economic cyberneticist, and many others. His ideas also
found early favor with Khrushchev, who tested the profit hypothesis in two
garment factories. Even after Khrushchev’s ouster in 1964, General Secretary
Leonid Brezhnev and Premier Aleksei Kosygin, an economic planner com-
mitted to systematic reform, continued to support most of Liberman’s ideas
in the partial and piecemeal roll out of the 1965 Kosygin-Liberman reform.21

9800.indb 65 6/2/16 3:05 PM

66  Chapter 2

The fundamental thrust of Libermanism, as it became known, was not

a sweeping reform of the command economy or its complex accounting
(for example, he retains several mandatory target measures in his 1962 arti-
cle) but rather a retooling and focus of command economy accounting on
profit—or what might be called a profit-in-command system.22 At the heart
of these reforms lay an attitude about information that other cyberneticist
economists and classical liberal economists on both sides of the cold war rec-
ognized at the time: it was an information index that reveals enough about
that product and its economic environment to be properly managed. For
free-market economists, that golden piece of information was the price of a
good; for Liberman, it was the profitability of an enterprise. This reform finds
its roots in a compromise between the preservation of the command econ-
omy administration and a sideways appeal to ongoing economic calculation
debates in Europe. Although Liberman could not explicitly argue against the
establishment of a central pricing board (as Friedrich Hayek did in 1945),
Liberman’s reforms appealed to the efficiency of decentralized economic
mechanisms that communicated local knowledge in real time without direct
administrative intervention. To Liberman in the late 1950s and early 1960s,
it appeared that a self-correcting marketplace of profitability might help
eliminate economic inefficiencies, if only factories and enterprises that gen-
erated more values than costs could receive their rewards.
Although these two indices—price and profitability—appear to stand
as key indicators that distinguish between liberal economists within and
without the Soviet Union for streamlining accounting problems besetting
any national economy, the opponents to Liberman’s reforms insisted that
reforming profit measures would also compel a concomitant reform in
price: for profit to be a meaningful index, it had to reflect relative scarcities
in the economy. This would make visible the hidden subsidies that the state
used in the existing pricing system to redistribute resources from one sector
to another. It is not clear that Khrushchev understood the full consequences
of his decisions: his statements on investment priorities were unclear and
changing, perhaps deliberately so, because as a staunch supporter of heavy
industry, he enjoyed the discretion to redirect and subsidize certain sec-
tors over others—the very discretion that full profit reforms would have
threatened.23 Nonetheless, the opponents to Libermanism—including the
cyberneticists—insisted that, whether in a market or planned economy, all
indicators were complexly interconnected. Changing one would surely pre-
cipitate a change in the other.
Liberman’s reforms met an uncertain end at the hands of those institu-
tions that implemented them in the late 1960s (simultaneously with efforts

9800.indb 66 6/2/16 3:05 PM

Economic Cybernetics and Its Limits  67

to advance the OGAS Project for economic reform). Adopted by Aleksei

Kosygin and implemented incrementally and partially by a hesitant new
general secretary, Leonid Brezhnev, the Liberman reforms nonetheless cor-
related with increased national production during the next five-year plan
(1965–1970), even though they also met fierce resistance from bureaucrats
and economic planners, especially in Ministry of Finance, who were set on
disrupting the raw materials supply chain and decrying the wage-differen-
tiating reforms as a form of class warfare.24 By the early 1970s, Brezhnev
continued to resist the orthodox economic planners but also abandoned
the Liberman reforms.
During the early 1960s, a third camp of thought about national eco-
nomic reform began to coalesce. The economic cyberneticists championed
what might be called planometrics, or a combination and application of
econometric mathematical tools that included input-output models (not
dissimilar from planned supply and demand), linear programming, and
sophisticated statistics to the problem of economic planning. Like the lib-
eral reforms, the economic mathematicians, cyberneticists, and economet-
rists comprising this loose camp conceived of the command economy as
a vast information-coordination problem. Unlike the liberal economists,
however, the cyberneticists were less concerned with reducing the com-
plexity of the economy understood as an information system to a single
golden index. They held that the other two camps did not take seriously
enough the numerical nature of all economic exchange and the capacity of
modern computing to process them. Mathematicians and theorists such as
Leonid Kantorovich, Vasily Nemchinov, Viktor Novozhilov, and B. Mikha-
levsky and in the mid-1950s cyberneticists such as Viktor Glushkov and
Nikolai Fedorenko realized that universal economic computability meant
that all economic relations could be modeled, optimized, and managed
with sufficient help from computers and their numerate keepers. In theory,
it did not matter which indices were considered, whether price or profit or
some proxy variable for peace or propaganda, so long as the boldest social-
ist ambitions for national economic and social justice could be calculated.
In theory, very fast computational speeds made this possible. Computers
were thus yoked, quoting Aksel’ Berg’s book series title on cybernetics, “in
the service of Communism” with more enthusiasm than any other tool-
kit before. By cutting through the political debates of the orthodox and
liberal economists, the cybernetists effectively intoned in the face of any
economic problem the immortal words of the patron saint of cybernetics,
Gottlob Liebniz in 1685: “calculemus” or “let us calculate, without further
ado, and see who is right.”25

9800.indb 67 6/2/16 3:05 PM

68  Chapter 2

The most prominent pioneer and precursor to economic cybernetics was

Leonid Kantorovich (1912–1986), a prodigious polymath who contributed
to the fields of mathematics, economics, and computer architecture. Kan-
torovich has been compared to John von Neumann (1903–1957), another
polymath born of middle-class Jewish parents in early twentieth-century
eastern Europe to contribute to the same fields (figure 2.1).26 Kantorovich’s
work on computationally optimizing economic exchanges, which later
became known as linear modeling, began before World War II.27 The only
Soviet economist to be awarded a Nobel Prize (1975), Kantorovich developed
linear modeling in 1939 to balance a series of competing variables algorith-
mically. A simple example adds a dash of dust bowl empiricism to its compu-
tational merger of both profit and planning logics. Suppose that farmers—or
after Stalin in the 1930s, the managers of a collectivized farm—distribute
crops across their fields and that the farmers know the cost of fertilizer and
pesticide, the cost of planting, and the selling price of wheat and barley. A
linear programmer can determine how much land they should devote to
each to optimize their annual yield. Linear modeling—now evolved into the
field of linear programming—allows the farm managers to calculate in matrix
form the maximum revenue, or profit, that they can expect from their avail-
able resources and to know how best to distribute their crops (for example,
how much barley and how much wheat to plant).28

Figure 2.1
Leonid Kantorovich

9800.indb 68 6/2/16 3:05 PM

Economic Cybernetics and Its Limits  69

Economists worldwide recognized the promise of this profit-by-plan-

ning model, especially after Kantorovich in 1939 and George Dantzig in
1947 separately took pains to propose methods that could scale to much
larger problem sets. Dantzig, for example, showed how the task of dis-
tributing seventy jobs to seventy people could be optimized, and Kanto-
rovich’s methods found aggregate use in the national wartime efforts to
maximize the costs of enemy losses and to minimize those of the Soviet
army. (Decades later, their methods remain in use today in modern opera-
tions research, such as in Walmart’s supply chain.)
Despite the apparent promise of profit by planning in military contexts,
linear modeling did not spread in Soviet circles after Stalin had dismissed
input-output “balances” as a “numbers game” in 1929.29 Sped by cybernet-
ics of the late 1950s and the translation into Russian of two articles (a 1958
translation of Wassily Leontief’s 1953 edited volume Studies in the Structure
of the American Economy and an article by Oskar Lange), the majority oppo-
sition to economic cybernetic planning methods in 1956 had become a
minority position by 1960, and momentum continued to build into the
late 1960s.30 By 1967, the Council on Cybernetics reported over five hun-
dred institutes and tens of thousands of researchers working on cybernetic
problems, over half of which featured economic cybernetic research. To
this day, the label of economic cybernetics lies exclusively within the former
Soviet Union and its area of influence.
The scaling successes of economic cybernetics in the late 1950s sug-
gested to Anatoly Kitov, Vasily Nemchinov, Viktor Glushkov, and others
that economic planning methods should be applied nationally—perhaps
even, as Kitov advised, in a real-time network of computers. The promise
of the scalability of the linear programming and computational methods
bolstered the political appeal of the supposedly apolitical planometric cal-
culation. The next step with a scalable computational tool is to scale it
all the way up, and that would require a communication infrastructure—
computer networks—for processing the nation’s economic coordination
problems. Because computational methods do scale, the economic cyber-
neticists enthused that maybe the principal question for economic reform
(who should control the command economy and how?) might be resolved
without either the price of politics of the politics of price. It might, the
cyberneticists reasoned, be solved with computers.
Many of these proposed reforms—cybernetic, liberal profit, and the Tay-
lorist reforms in the 1920s under Lenin—merited serious attention and, if
implemented, would likely have borne fruit had they not collided in appli-
cation with serious institutional constraints from the bureaucracy. It was

9800.indb 69 6/2/16 3:05 PM

70  Chapter 2

self-evident to the economic bureaucracy that computers were not value-

neutral: cyberneticists ran them, and no state resolution could convince the
bureaucrats to behave like rational bureaucrats in ceding power to cyber-
neticists. The resulting messy resistance and nonhierarchical dynamics of
the administrative base that directed the Soviet command economy reveal
institutional tensions and contradictions that foreclosed against multiple
attempts to reform the national economy computationally, liberally, and
otherwise. Just as Khrushchev’s reforms were frustrated and fractured by the
internal resistance of administrators who clung to the current positions of
power in the late 1950s and early 1960s, so too did the cybernetic appeals to
technocratic reform begin to break against the practical problem of reform-
ing a national economy that refused to behave like the hierarchical system
that it appeared to be on paper.
Liberal economists and economic cyberneticists (at least initially, under
Viktor Nemchniov) appeared to be bound for a great alliance. In the early
1960s, Nemchinov proposed a “self-supporting (self-accounting) system of
planning” that integrated both decentralized computational and market
mechanisms into the planning apparatus. The basic proposal was to solve
the incentive problem in a way that no factory would have a reason to
act against the wishes of the center and the center would have no reason
to compel the factory to act.31 With time, however, the cybernetic econo-
mists and the liberal economists clashed over whose method would win
the balance of state approval. In 1963, both Liberman’s profit proposal
and Glushkov’s OGAS project appeared positioned to affect real economic
reforms. Leading liberal economists, including Evsei G. Liberman, A. M.
Birman, and B. D. Belkin, voiced the public opposition to mathematical
economic reform in general and the OGAS Project, in particular, although
without spelling out the secret project by name in the press. These lead-
ing liberal economists immigrated to the United States and Israel after the
Liberman-Kosygin reforms were formally accepted but botched (or rather
deliberately butchered) by the administrative apparatus. Birman criticized
the economic cyberneticists not for their methods but for their politics. As
late as 1978, he contended that the introduction of computers and auto-
mated systems of management (ASUs) into Soviet economics constituted
no more than a “costly delusion” and was a question of the complexities of
human interests, not precise accounting.32 In effect, the liberal economists
accused computational economics of harboring conservative politics and
of trying to work in the framework of the existing political system without
any social change.

9800.indb 70 6/2/16 3:05 PM

Economic Cybernetics and Its Limits  71

Most cyberneticists came from the technical, theoretical, and natural sci-
ences—fields that attracted many of the brightest Soviet minds because of
the state support they received and the safety of choosing ostensibly non-
political specialties. The competition to join the top sciences was immense,
and few chose to leave the sciences for the social and humane sciences
(Kitov was forced into economics, and Glushkov was an exception). Lead-
ing figures in the mathematical economic camp (such as Kantorovich, who
received the Nobel Prize in 1975) were known to defend orthodox politi-
cal values about the price of labor, even while the younger generation of
cyberneticists sought to avoid the politics of price by arguing that a suf-
ficient change in the organizational values of the system must also cause
a concomitant change in the political values. By attempting to rationalize
and decentralize the planning process, the cyberneticists hoped that any-
one, with the help of a computer, could contribute to a reformed, well-oiled
economic model and plan, make the system work better, and open a quiet
back door to political reform. Even so, Birman and other veteran economic
reformers wondered whether the deliberate planning that was inherent in
a cybernetic reorganization of economic planning would exacerbate and
reaffirm preexisting constraints and coordination problems in the com-
mand economy. The liberal economists saw in cybernetic reform of the
planning administration no promise of a transition to the market economy
that they sought. This belief that technological and organizational reforms
bring political ramifications recurs as an article of faith in the annals of
Soviet cybernetics.
Despite Glushkov’s complaints to the contrary, it is not clear that lib-
eral economic opposition to the economic cybernetic school held up or
delayed Soviet attempts to carry out economic reform by computer net-
works.33 By 1970, when the top echelons of the Party were ready to consider
such proposals in earnest, the liberal economic opposition to the cyberneti-
cists might have helped ingratiated the cybernetic cause to more orthodox
Party members who were fed up with Libermanism. (By that point, liberal
reforms had a five-year track record of generating more heat than light in
many of the economic administrations.) At the same time, the military and
the Party were tantalized by the promise of a third generation of integrated
circuits in computing in 1970s and maybe even the fourth generation of
microprocessors on the horizon of computing industries abroad. Given the
political and technological climate, the ears of the state were primed to hear
Glushkov’s declaration that “the scientific-technical revolution has thrown
such a challenge to the science of governance, and much will depend on
how we dare to answer that challenge.”34

9800.indb 71 6/2/16 3:05 PM

72  Chapter 2

Vertical Bargaining and Other Organizational Dissonance in the Soviet

Command Economy

This chapter’s consideration of the inner workings of the command econ-

omy looks directly into the political heart of socialist economic reform, of
which the cybernetworks were a small part. To organize an economy prop-
erly was the litmus test for the Soviet experiment in socialism and social
justice. Few other national projects claimed and endured as much in search
of a Soviet network. As a consequence, the organizational dissonance that
all economic reformers, not only cyberneticists, encountered in trying to
make the economic numbers line up was both the cause and effect of con-
tinual economic reform. In this industrialist mindset, computers brought to
perennial problems a new set of tools (linear processing, input-output mod-
eling, and the possibility of real-time network communication and surveil-
lance). This section examines some sources of what David Stark has called
the “organizational dissonance” that underlies the command economy and
that helped ensured the economic system could not be reformed or reaf-
firmed because every reform introduced new problems without solutions.35
“Vertical bargaining” was a feature, not a bug, of the perpetual misalign-
ment of incentives in the Soviet economic hierarchy. Named by a Hungarian
economist and critic of socialist economic systems, János Kornai, vertical
bargaining takes place among the three levels of relationships among a local
enterprise, a branch directorate, and the national planning ministry. Verti-
cal bargaining took place continuously in the annual planning process that,
as Spufford describes it, pulsed with paperwork between Gosplan, regional
councils (or Sovnarkhozy between 1957 and 1965), and the enterprises (such
as firms, factories, farms). Every spring, the enterprises asked Gosplan for the
supplies they needed as a percentage change from the output of the previ-
ous year. Around the end of June, Gosplan sent draft production targets to
the regional councils, which disaggregated the targets and then negotiated
with the enterprises toward trim but not unmanageable requests for inputs.
Gosplan then reaggregated these requests into each commodity’s total sup-
ply for the nation that year. When the figures did not match, a second nego-
tiation period between Gosplan and the regional councils proceeded into
the autumn until Gosplan had limited demand and maximized supply. The
finalized supply quotas and production targets could then be passed down
the chain in late October to allow enterprises to select next year’s items from
the “specified classification,” a list of every item that officially was produced
in the Soviet Union (think of the Sears mail order catalog on steroids, minus
the advertising), just in time for the process to begin again.36

9800.indb 72 6/2/16 3:05 PM

Economic Cybernetics and Its Limits  73

All negotiation processes were structured for expressing “mutually con-

tradictory motives,” although the administrators had no special access to
mechanisms for resolving a priori conflicts between the interests of the lay-
ers of the hierarchy, which compromised the integrity of the economic plan
that they were developing.37 Without a plan for regulating the planners, the
planning processes confronted economic leaders—everyone from adminis-
trative planners to factory managers—with multiple registers of conflicting
value. “Suppose a leader feels he has received an incorrect order,” Kornai
asks: “Should he carry it out or should he protest, out of party loyalty and
professional pride?”38 If he accepts the flawed order but fails to deliver on
it, Kornai continues, he and his colleagues will be held responsible and pos-
sibly accused of sabotage. If he opposes the order, he could be accused of
party disloyalty. Either way, the actor, not unlike Vanek in Vaclav Havel’s
play Audience, is stuck.
Without a single path forward, levels had to negotiate for their own
institutional self-interests vertically across the formal administrative hier-
archy. To do so, requests began to misrepresent economic reality in both
directions. Requests for input (or demand) rose upward and request for
outputs (or supply) sank downward—the planner’s vertical equivalent of
selling high and buying low in a horizontal market. Imagine the behavior
of the ministry that oversees a branch directorate and the factories that
the directorate oversees. The branch directorate is charged with reporting
to the ministry statistics about the annual production, material allocation,
and labor of its subordinate factories. To do so and because the experienced
directorate anticipates that factory managers are responsible for shortfalls
and thus systematically underestimate their output capacity and overes-
timate their input needs, the director will “prescript a plan 10 or 20 per-
cent tighter than they themselves consider realistic, calculating that the
firm will want to beat them down.”39 When reporting its plan, the branch
directorate bids to the superior ministry just as the factory manager did to
it. An apocryphal anecdote of a job interview for a new accountant in a
factory captures something of this haggling spirit. To each candidate, the
factory manager asks only one question: “How much is two and two?” A
single candidate, a former convict, has the winning answer: after hearing
the question, he stands, closes the door, and asks in a loud whisper, “How
many do you need?”
The vertical bargaining process also penalized the future of productive
factories by “planning in” their previous successes as the new baseline,
ensuring that the plan would be ratcheted upward in perpetuity.40 Man-
uel Castells notes that the entrepreneurial managers and workers in the

9800.indb 73 6/2/16 3:05 PM

74  Chapter 2

chemical complex of Shchenkino in Tula, Russia, “were trapped into being,

in fact, punished with an intensification of their work pace while firms
that had kept a steady, customary level of production were left alone in
their bureaucratic routine.”41 This ratcheting effect, a kind of institutional-
ized variant of the tall poppy syndrome, has its corollaries in the mutu-
ally reinforcing relationship between demand and supply. In corporate and
command regimes, if the supply of one’s quality goods meets demand, one
must work harder to meet future elevated demand. If the supply of one’s
goods falls short of the need, the capitalist market actor will adjust or go
bankrupt, and the socialist administrative actor will be punished. So long
as labor is isolated from those who manage the means of production—Marx
himself railed against the doyens of exchange value (Tauschwert)—manage-
ment profits and alternately pays or punishes workers for past productivity.
Unlike market behavior, any deviations from the plan could send culpable
ripple effects down or up the chain, and the plan itself could be understood
in the context of its local knowledge. So the ideal standard of factory or
firm behavior is to fulfill the plan by “exactly 100 percent, or perhaps 101
or 102 percent.” The lively fitfulness of vertical bargaining disrupted and
distorted the representativeness of the economic statistics that the cyber-
neticists sought to input into the linear modeling and programming and
economic reform projects.
Shoehorned instead into numerical fit, the command economy plan
never fully reconciled in its manifold details, and the compounding compli-
cations and activities that organizational dissonance invited were anything
but planned. Administrative roles blurred as needed. One leader might find
himself playing the politician, a bureaucrat, a technocrat, and a manager
in different situations. To secure more input on the accounting sheets, a
leader might encourage workers on the factory floor to produce more out-
put in the name of the plan but, behind closed doors, might misrepresent
the numbers to undercut the same plan. Not at all the rigid hierarchy cari-
catured in modern memory, the everyday practices of Soviet economic life
abounded in pervasive informal forms of competition and caprice.
These early cybernetic economists faced an associated and monumental
challenge. In theory, their reforms would be approved to track the formal
(or white or first) economy, and any attempt to reform the distortions in
that formal economy would further incentivize more informal (or gray, sec-
ond, or shadow) economic activity. The reality of the latter half of Soviet
economic history reveals that private life increasingly depended on the
resilience and robustness of people’s connections to the informal economy.

9800.indb 74 6/2/16 3:05 PM

Economic Cybernetics and Its Limits  75

The command economy operated on hidden networks of tolkachy (liter-

ally “pushers”) or “go-to-guys” or “fixers” who got the job done outside
of the formal economic plan. Without the support of tolkachy, thousands
of official economic quotas over decades never would have been met.42
Recent economic studies based on previously unconsulted archival data
have estimated that a stunning average of 24 percent of annual expendi-
tures per household in the Soviet Union went to the informal economy
between 1968 and 1990.43 The estimated percentage of GNP not accounted
for by the informal economy over the same period ranges between 17 and
40 percent.44 Despite both official claims and its enemies’ fears otherwise,
Soviet economic life drew its vitality not from the strictures of top-down
command and control but from the fitful hustling and the scrambling that
came about because of those commands.
Informal behavior and bargaining were not separate from Soviet state-
craft: the state embodied them. Even Stalin, with his reputation as an all-
knowing leader and steely strongman, bowed to the deeper logic of informal
influence and favors—what is known simply as blat (ostensibly from Polish
Yiddish for “someone who covers for someone else,” or from the German
for “blank note”).45 Instead of committee decisions, Stalin often invited
local leaders to private consultations where Stalin could claim that all other
parties had endorsed his recommended policy and provide the local del-
egates with an opportunity to leverage “personal connections” to personal
advantage at home.46 Control over science and society was extended by the
same informal means. In 1952, an editorial in Hungary proclaimed that
“the teaching of Stalin embraces all the universal principles of nature in
its smallest details. He solves all the practical problems of understanding
natural science,” and “it is only Stalin … who is able to analyze clearly and
find with mathematical precision the exact way toward solution of present
day problems.”47 So too did his social radar appear impossibly omniscient
thanks to strategically placed ambassadors and Party secret police embed-
ded with party apparatchiks. The strongman seemingly did not want the
state to behave as a well-ordered hierarchy but rather as a sprawling net-
work with informal connections to a strong but sporadic center. The terror
of his rule was not its rigid centrality but its informal uncertainty. Stalin
and his henchmen could be anywhere. To encompass everything was Sta-
lin’s job: he already had everything covered. And for this reason, his death
formed the vacuum into which cybernetics stepped.
Khrushchev’s thaw attempted to distance the nation from its Stalin-
ist past, but his decentralizing reforms sped the sporadic, informal, and

9800.indb 75 6/2/16 3:05 PM

76  Chapter 2

unregulated nature of his rule. Even his famous “secret speech” to a crowded
Party congress in 1956, which was the act that distanced him from Stalin,
denounced Stalin as a cultish personality but not as a governor whose mode
of informal management Khrushchev wanted to break from.
The administrative infrastructure came to reflect this infusion of infor-
mal administration in a number of ways. Administrative personnel and
staff officials had separate telephone lines and mailboxes for the same
supervisor, which ensured that formal communication lines were clogged
with official requests and that the actual negotiations took place along
informal lines—not on the golf courses of modern business but in the tran-
sit sites such as hallways, trains, and dachas (seasonal cottages or summer
homes outside the city). Because formal mechanisms proved ineffective,
hiring and promotion practices often relied on interpersonal and informal
“career friendships” or tight bonds that lasted lifetimes. Soviet specialist
David Granick notes that “with this absence of formal clarity, it is natural
that emphasis has always been placed on the need for the closest ties and a
comradely atmosphere between the management and a plant’s Party orga-
nization.”48 Interviews with émigré bureaucrats have revealed a pattern of
administrative behavior that stressed the career necessity of not “spoiling
relations,” the significance of who you know, and the career advantages
of being a “yes-man” in formal relations with superiors.49 Administrative
conflicts between the elements of that system—such as the Academy of Sci-
ences and the Ministries of Finance, State Planning, Interior, Defense, and
State Security (home of the KGB, or Committee for State Security)—were
resolved not by an appeal to hierarchical authority but through a variety of
informal mechanisms that were internal to the ministries themselves. One
reform initiative after another was aborted, and those that were enacted
were condemned to stumble on, in Schroeder’s phrase, the late Soviet
“treadmill of reforms.”50
Compelled to operate within an official hierarchy, Soviet administrators
benefited by behaving and working across complex informal networks that
crisscrossed across institutional interests. A young economist, Menshikov,
summarized the nonlinear or nonhierarchical behavior of the command
economy not as “aiming at increasing the well-being of the population”
but as “maximizing the power of the ministries in their struggle to divide
up the excessively centralized material, financial, labour, natural, and intel-
lectual resources.”51 He continues, noting the path-dependent creep of
administrative misbehavior: “Our economic-mathematical analysis showed
that the system had an inexorable inertia of its own and was bound to
grow more and more inefficient.” Whether in vertical bargaining between

9800.indb 76 6/2/16 3:05 PM

Economic Cybernetics and Its Limits  77

the levels of the planning, the reluctance of ministries to cooperate out-

side their assigned territory, or the struggle of other agencies to collaborate,
the fundamental paradox of planning became increasingly obvious to the
economic cyberneticists and others who sought systemwide reform. The
planners and executers of that plan were to create gaps in the plan and
leverage those gaps with very competitive logics that the command econ-
omy sought to prevent.
The cyberneticists thus faced a foundational paradox in reforming the
national economy and perhaps other political economic systems. For an
economic reform project to succeed politically at the national level, the
reformers had to win the support of the very system that it meant to reform.
If they sought to do so through new formal mechanisms, as the computa-
tional methods of the economic cyberneticists demanded in theory, those
methods would face widespread resistance (one of few systematic behaviors
that the system was regularly capable of). Conversely, if they sought to
reform a broken system of political patronage, as they had to do in practice,
they first had to win the favor of that system. The paradox that they faced
is not unique to the Soviet cyberneticists. To reform a system, a would-be
reformer first has to become part of it. Next, the better that one plays along,
the less likely that one wants to reform the system; and so long as one con-
tinues to play along, one may not reform the system.
Faced with technocratic reform, economic management bureaucrats
and politicians scrambled their own administrative orders to preserve their
own personal careers. Bureaucrats were never mere bureaucrats, and the
mechanics of day-to-day operations were never merely mechanical, even
though the culture of technocratic governance swelled after Stalin to the
point that, by 1989, 89 percent of those who sat on the Politburo were
trained engineers (engineering training prepared Soviets for governance
positions much like law degrees do elsewhere).52 (The iconoclast economist
Thorsten Veblen mused in 1921 that the West might one day be ruled by a
“soviet of technicians” or a technical class that was capable of capturing the
wealth that they produced.53) The Soviet system, much like a firm, sought
to produce one solitary good above all else—the political good of a life apart
from the capitalist experience. In a narrow sense, it succeeded: the mar-
ketplace of Soviet economic interactions became foremost a negotiation
of political power rather than price. Its bureaucrats bowed to unintended
incentives to exploit the rampant organizational dissonance that they over-
saw, its technocrats lived by their social wits, and the system squeaked by
on the capricious politics of planning run amok.

9800.indb 77 6/2/16 3:05 PM

78  Chapter 2


As Soviet economic cyberneticists emerged as a viable option in the early

1960s and again early 1970s, they confronted a monumental problem in
managing and reforming the command economy. The question was identi-
fied by the Austrian school of economics decades earlier: which techniques
and approaches would help resolve the mounting tensions among the for-
mal command economy, the gray economy, and the infusion of informal
practices in the administration of Soviet socioeconomic life? For the most
part, the leading Soviet economic cyberneticists sought to fix the formal
command economy by introducing ambitious, even grandiose, plans for
automating and modeling the administrative planning decision process
itself. And yet, as is shown, those formal plans—a networked plan to fix the
planning process itself—did not work because even cybernetic plans could
not account for nonlinear operations in the Soviet economy. Their formal
plans to rebuild the command economy as a hierarchy had to overlook
the complex crisscrossing networks of relations that made it function in
practice—the gray economy and its entrenched currency of blat or informal
favors that were entrenched in the governance structures. By reimagining
the command economy as a heterarchical crisscross of hierarchical orders
from above and a resulting swirl of unregulated practices in every other
direction, the failure of these Soviet economic cyberneticists to reform,
automate, and manage the command economy begins to make more sense.
There is one goliath exception to this critical description of the infor-
mal administration of the state and economic bureaucracy. The command
economy, which staggered along a winding path toward the creation of a
normal industrial civilian economy, was relatively functional at powering
and sustaining superpower military technological initiatives. Formulated
first as a wartime economic model by the Germans, the insatiable sink of
the Soviet defense apparatus into which economic resources were continu-
ously poured cannot be overestimated. Both official state and CIA statistics
on Soviet military spending are controversial, although if a critic of both
can believed, the Russian American economist Igor Birman estimated that
by 1975 the CIA estimates of the size of the Soviet economy were two or
three times larger than reality and that instead of spending roughly 6 per-
cent of its GDP on military expenditures, the Soviet state devoted closer
to 30 percent of its GDP on the military. The military-industrial complex
enjoyed massive funding streams and the brightest and best intellectual
and technological resources, and although the jury is still out on the exact
nature of the Soviet military (most of its details remain closed to this day),

9800.indb 78 6/2/16 3:05 PM

Economic Cybernetics and Its Limits  79

the military sectors also attracted the best and the brightest because those
sectors were best managed. The strictly managed military sectors produced
and sustained for decades world-class space and nuclear programs and
secret computer networks across launch pads deep into Siberia. But the
Soviet military’s technological innovations did not as a rule spill over into
civilian sectors. Nuclear-blast-resistant computer chips interested very few,
and yet the Soviet military nonetheless developed these and other “special
research” projects (so named in public documents).54 The military enjoyed
the status of being responsible for generating a seemingly infinitely defensi-
ble “public good”—national defensive and offensive readiness in an almost
irrationally strategic cold war—and yet did not have the burden of having
to be publicly accountable to civilian politicians.55 Perhaps the caricature of
the problems of the civilian economy makes most sense in light of its foil
in the military economy. Unruly, informal, labyrinthine, and ineffectual
suffering in the civilian sectors rarely met with the well-ordered, formal,
hierarchical modernization in military affairs. The contrast between mili-
tary and civilian economies recapitulates a structurally similar disconnect
between the civilian command economy in theory (hierarchical, formal,
well ordered) and in practice (heterarchical, informal, conflict ridden).
In summary, the separation between military and civilian sectors
reflected a disconnect between the civilian economy and its own state
goals, and it comes squarely into play in the central story, outlined in the
following chapters, about early Soviet computer networks projects and
those cybernetic entrepreneurs who set out to build them. This chapter
has laid out the basic civilian economic operations as well as problems
that motivated Soviet cyberneticists—whether orthodox, liberal, or cyber-
netic—to propose and design ambitious projects for reforming the national
economy. The next two chapters examine the role that computer networks
played as the promised deliverers of such reform, and they turn on why
cybernetic attempts to network the command economy fell apart. The net-
work entrepreneurs understood firsthand the institutional contradictions
that they sought to solve with an automated system of management. This
basic backdrop to the everyday administrative conflicts in Soviet social
life—between the rational hierarchical plan of the command economy and
its messy heterarchical misbehavior—was not lost on the pioneering Soviet
cyberneticists who followed.

9800.indb 79 6/2/16 3:05 PM

9800.indb 80 6/2/16 3:05 PM
3  From Network to Patchwork: Three Pioneering Network
Projects That Didn’t, 1959 to 1962


In the late 1950s and early 1960s, economic cybernetics—with its nonlinear
mathematical mindset—appeared to be a near-perfect approach for mod-
eling and reforming the economy’s heterarchical coordination problems.
(Cyberneticists were at home in the underlying observer-effect problems: From
the act of looking at a system changes the system.) Between 1959 and 1982,
Soviet cyberneticists advanced a half dozen ambitious and creative plans
to network their nation, including four overlapping attempts to digitize
the command economy. Most of those proposals arose between 1959 and
1962 before they languished or merged with associated projects with more
momentum, like the OGAS Project described in chapters 4 and 5.
This chapter reviews the three earliest Soviet network ambitions—Ana-
toly Kitov’s EASU (Economic Automatic Management System), Aleksandr
Kharkevich’s ESS (Unified Communication System), and N. I. Kovalev’s
rational system of economic control between 1959 and 1962—in the con-
text of the larger institutional struggles to secure support for their network
projects. Some attention also is paid to the network projects that developed
as civilian computer networks elsewhere and to the ways that the Soviet
experience varies from the American ARPANET and Chilean Cybersyn proj-
ects. Through a discussion of leading Soviet network proposals to reform
the command economy in the last few years of Khrushchev’s reign (1959–
1962), this chapter examines, details, and complicates the hypothesis that
the administrative dynamics of a strong civilian-military separation help
explain the stillbirth of their historic efforts.

Anatoly Kitov and EASU: The First Soviet Cyberneticist and His
Civilian-Use Military Network

The first person to propose a large-scale computer network for civilian

use anywhere, as far as I can tell, appears to have been the first Soviet

9800.indb 81 6/2/16 3:05 PM

82  Chapter 3

cyberneticist, Anatoly Ivanovich Kitov (1920–2005). The son of a White

army (Menshevik) officer who escaped persecution after the 1917 Russian
revolution by moving from Moscow to central Asia and then to the city of
Kyibishev (now Samara, Russia) on the Volga River, Anatoly Kitov grew up
between two world wars. A star student in mathematics who rose rapidly
through the military academy, Kitov served as a young officer on the front
in World War II (where he, like other human “computers,” computed bal-
listic tables), before launching a distinguished military career that suddenly
shifted to civilian research for reasons described below (figure 3.1).
In 1953, Aksel’ Berg, then deputy minister of defense in charge of radar
and future dean of Soviet cybernetics, asked Kitov to prepare a report on
the state of computing in the West.1 Kitov’s optimistic report resulted in the
creation of three large computational facilities—the Computation Center
1 (which Kitov directed until 1959), the Navy Computation Center, and
the Air Force Computation Center.2 Kitov’s optimistic review of comput-
ing in the West stemmed from his 1952 discovery of a copy of Norbert
Wiener’s Cybernetics that had been removed from general circulation (due
to the ongoing anti-American campaign against cybernetics) and stored in
a top-secret military research library. As noted above, in 1955, Kitov coau-
thored (with Lyapunov and Sobolev, two highly regarded Soviet mathema-
ticians) the first Soviet article to attempt to rehabilitate cybernetics from

Figure 3.1
Anatoly Kitov. Courtesy of Vladimir Kitov.

9800.indb 82 6/2/16 3:05 PM

From Network to Patchwork  83

the anti-American ideological critique that had been waged since its first
mention in the Soviet press in 1948.
Kitov was not alone in seeing the potential for using computers in mili-
tary work. Military and computing innovations were inseparable in the early
history of computing. Although those early, specialized computer innova-
tions for the military often had no measurable defense outcomes, their
technological innovations seeped into nonmilitary industries. (Examples
of military products that are now available commercially include jet planes,
semiconductors, telecommunication and computer equipment, microelec-
tronics, sensors, GPS, drones, and even Velcro, and only a few consumer
electronics, such as game consoles and consumer electronics, have run the
other way.3) In the United States during World War II, early military com-
puter projects included the Whirlwind I (a vacuum tube computer), the
Whirlwind II, and early attempts at computerized command and control,
which in the 1950s led to SAGE (Semi-Automatic Ground Environment),
which was a radio and radar network that stretched over most of Canada
and was intended to intercept invading bombers from the Soviet Union.
For its military purposes, SAGE was obsolete before it was operational, but
its preparation nonetheless sparked a wave of influential inventions and
major technical advances in computer systems and networks, including
magnetic core memory, video displays, graphic display techniques, analog-
to-digital and digital-to-analog conversion, multiprocessing, and automatic
data exchange among computers.4
Alarmed by the news of SAGE in the mid-1950s, the Soviet military
responded with at least three major long-distance computer networks—a
missile defense system (System A in the late 1950s), an air defense system
(the TETIVA in the early 1960s), and a space surveillance system (begin-
ning in 1962). In the late 1950s, for example, a prototype missile defense
system that was code-named System A (about which little else is known
today) was built around a computer network that connected two Soviet
mainframes, the M-40 and M-50, and a series of specialized computers at
remote radar installations. Soon after the successful testing of System A in
March 1961, Khrushchev boasted that Soviet antiballistic missiles could, in
his famous phrase, “hit a fly in outer space.”5 More significant than the sys-
tem’s accuracy, however, was the fact that System A and its sibling military
networks compelled a larger geostrategic shift: namely, antiballistic missiles
that pointed skyward around the world greatly diminished the strategic
value of a first-strike attack. Other networks, such as the space surveillance
system started in 1962, connected a pair of distant nodes (one near Irkutsk
on Lake Baikal in eastern Siberia and the other in Sary-Shagan in south

9800.indb 83 6/2/16 3:05 PM

84  Chapter 3

central Kazakhstan) to a master computer center just outside of Moscow

some seven thousand kilometers away. Other radar networks—named Hen
House, Dog House, and Cat House—bathed vast swaths of territory in anti-
ballistic alerts. The Dead Hand—the semiautomatic perimeter defense sys-
tem noted in the introduction—is the present-day heir to these early Soviet
military computer networks.6
Struck by the latent computational surplus that was available in mili-
tary networks, Kitov turned his attention to how networked computing
might be able to benefit the civilians who were supposed to be protected
by the military.7 In 1956, in the first Soviet book on computers, Digital
Computing Machines, Kitov built on the insights of Leonid Kantorovich’s
linear modeling to promote a pioneering argument about the potential of
the computer—although still without the network—as an essential tool for
modeling, programming, and regulating the Soviet economy.8 The idea of
using high-speed digital computers to crunch economic statistics was noth-
ing new. In fact, Kitov’s proposal came as a mere technological update to a
long-standing tradition of state-based mechanical computation. After the
1917 revolution, for example, the Bolsheviks quickly nationalized the Odh-
ner calculator factory in St. Petersburg. By 1929, the year that farms were
mass collectivized and planned, the Soviet Union employed statistical tabu-
lation equipment—including a clone of the successful pinwheel Odhner
arithmometer and IBM punch card machines—on the scale of the United
States or Germany.9 The etymology of the word statistics (German for “the
science of the state,” which replaced “political arithmetic” in English in
the late eighteenth century) frames this trend as nothing peculiarly Soviet:
modern states, from the census to taxation and conscription, have long
made statistics their business.10
The initial idea of using unnetworked computers to process economic
information—really no more than a modest technological upgrade that was
in line with well-established state interests—had been gathering support
since at least 1954, when Mikhail Kartsev, an engineer who participated
in the construction of the M-1 computer, declared that cybernetic under-
standing of computers had to exceed narrow military tasks: “we are inter-
ested not so much in the military applications of mathematical machines
or, more generally, new technical devices, but in their wider applications.”
His colleague Nikolai Matiukhin, citing the use of computers in U.S. busi-
ness, stressed that “in a socialist country, … the mechanization of planning
with the assistance of computers can and should be pursued to the largest
extent possible.”11 The early Soviet information technologist Isaak Bruk,
who developed the M-2 computer, picked up on the thread, publishing

9800.indb 84 6/2/16 3:05 PM

From Network to Patchwork  85

the first call to harness computational power into raising the quality of
economic planning in a 1957 Kommunist article, “Electronic Calculating
Machines: In the Service of the National Economy.”12 (Both Kartsev and
Matiukhin continued serving in military careers.)
Not long after, from 1958 to 1959, Kitov and Bruk’s proposal began
to bear fruit as Gosplan constructed specialized computational centers
(vyichislite’nyi tsentri) for economic accounting, which were to be under the
control of the Economic Council within the Council of Ministers. Vasily
Nemchinov, a leading Soviet mathematical economist, also championed
the proposal for computer centers around the nation to improve planning.
In January 1959, Kitov sent his first proposal directly to the top of Soviet
power and called on General (then First) Secretary Nikita Khrushchev to
recognize the need to use computers to process economic planning and to
speed economic reforms, and with the letter he included a copy of his pub-
lished book on digital computers.13 None of these proposals mentioned a
computer network—or its near synonyms (such as base, system, or complex)—
that would be able to command the economy.
Kitov’s first letter to the Central Committee in 1959 proved to be a
success. Although Khrushchev probably never saw the letter, his message
ended up in the hands of Leonid Brezhnev, who replaced Khrushchev as
general secretary in 1964. A trained technologist who had studied metal-
lurgical engineering in eastern Ukraine, Brezhnev approved the proposal
and ordered a government resolution to review and execute Kitov’s recom-
mendations. A commission led by Aksel’ Ivar Berg was created to enact the
Communist Party resolution called “The Speeding and Widening of the
Production of Calculation Machines and Their Application to the National
Economy.” Berg cut a cosmopolitan figure (Berg’s mother was Italian, and
he was a Swedish Finn) in the administrative support of Soviet cybernet-
ics, and he rose through the military ranks to serve on the State Commit-
tee of Defense after World War II with special emphasis on military and
naval matters. Rescued from a sunken submarine near Helsinki in 1918, he
directed radio technology research for the Red Navy, was imprisoned dur-
ing the Great Terror for spying, returned to research during World War II,
and was vice minister of defense in charge of national radar and radio tech-
nology from 1953 and 1957. In this position, Berg provided critical admin-
istrative support to the efforts of many early cyberneticists, including Kitov,
Lyapunov, and others. Appointed to serve as the chair of the Council for
Cybernetics in 1959, Berg, perhaps not unlike Vannevar Bush, exerted an
extensive range of administrative influence over the state of Soviet cyber-
netic science for the next two decades.14 With Berg’s support, Kitov’s first

9800.indb 85 6/2/16 3:05 PM

86  Chapter 3

letter to Khrushchev set into motion a sea change that swept up strong state
support for cybernetics.
A signal political victory, Kitov’s initial vision imagined computers as
devices for local computation but not yet for national communication.
Like his predecessors, he proposed that “electronic calculating machines”
must be used in “automating administrative and economic governance”
in planning for the then seven-year plan for the command economy.15 He
also called for a “reduction of the administrative-management personnel”
for engaging in “outdated means and methods of leadership.” This could
take place at local levels by means of local control systems, called automated
systems of management (ASUs) (avtomatizirovannie sistemi upravleniya). ASUs
were automated control systems at the level of the factory—a kind of local
area network that allowed mainframe computers to control and communi-
cate with factory machinery through a series of automated feedback loops
and programmable control processes. In the 1960s, ASUs were developed
and implemented in individual Soviet factories incrementally, and their use
increased slowly in the 1970s and 1980s.16 By the early 1960s, the notion of
ASUs—or computer systems for monitoring local industrial processes and
automatically optimizing those processes for efficient outcomes—gained
popular traction among factory and enterprise managers around the coun-
try. On the cusp of the 1960s, enterprising military researchers like Kitov saw
the advanced computer technology behind ASUs as promising new efficien-
cies, savings, and economies of scales at the factory and enterprise levels.
The next step was to nationalize the ASU and make it go “all-state” (add-
ing the prefix OG for obshche-gosudarstvennaya to form the OGASU). This step
appears obvious today, but the consequences of that step must have been
hard to foresee then. In fact, encouraged by the success of his first letter,
Kitov shifted his attention from local computation to national communica-
tion. In the fall of 1959, he drafted and sent to the Party leadership a second,
more ambitious letter that eventually became known as “the Red Book” letter
due to the color of its cover. The Red Book letter embraced a far more radical
idea—the first wide-area dual-purpose computer network that could support
both military and civilian uses. In his proposal, Kitov conceived of a “unified
automated computer network” for administrative control of both military
and economic affairs that would be built on the extant territorywide lattice
of Ministry of Defense computer centers. Although electrical telegraphs had
been instantaneously connecting paying publics across long distances since
the mid-nineteenth century, evidence indicates that Kitov’s 1959 proposal
for a dual-use network was the first anywhere to suggest allowing civilians to
use military computer networks to work on national problems.

9800.indb 86 6/2/16 3:05 PM

From Network to Patchwork  87

Kitov named this first national network the Economic Automatic Man-
agement Systems (EASU, for Ekonomicheskaya avtomatizirovannaya systema
upravleniya), a short step from the local area networks of the ASU. The EASU
was meant to be more than a colossally oversized ASU, or factory-based
automated management system, however: it was to be a dual-use network
that would overlay and manage existing information flows within the Soviet
economy with “large complexes” of computer centers. The long-distance
economic ASU began to articulate in technological terms the underlying
Marxist conception of the national economy as a single complex industrial
body. The informing values for the earliest Soviet networked vision were in
context both technologically ambitious and political self-evident.
The EASU proposed for the first time a long-distance communication
infrastructure to transform the command economy into what it had effec-
tively thought itself to be—a single nationwide corporation devoted to
producing one product, which was social life outside the reach of capi-
talism. The basic communications infrastructure for such an upgrade was
fairly straightforward. What Kitov called a “complex of computers,” or
a computer network, would use Ministry of Defense computers to opti-
mize national economic planning and streamline the bulky and inefficient
administration for planning the Soviet national economy. Each powerful
computer center in the network would build on military computing loca-
tions that already were underground, well protected from the threat of
enemy bombing and natural interference above ground. In addition, each
underground military computer center would connect to accessible com-
puter terminals that were located in cities above ground where “civilian
organizations” could receive, send, and employ “unlimited quantities of
reliable calculating processing power.”17 The military’s automated missile
computer network systems would serve, in Kitov’s vision, as the techni-
cal platform for computationally monitoring and managing the national
Kitov went considerably further than most military men of the time in
proposing dramatic financial savings and benefits for the state. Electronic
economic reform would also help quicken, Kitov added, the currently slug-
gish and inadequate adoption of computer technology by the Ministry of
Defense. As part of that criticism, he called for the creation of a new gov-
ernmental body that would be charged with overseeing the reform of all
institutions, including both military and civilian, that were associated with
planning the national economy.
With a proposal that criticized the military and proposed a civilian-mili-
tary project, Kitov sent his letter to Khrushchev sometime in the fall of 1959

9800.indb 87 6/2/16 3:05 PM

88  Chapter 3

with the “hope that it would be accepted,” he later remarked, “just as easily
as the previous one.”18 Unfortunately, the precise fate of Kitov’s second let-
ter—the Red Book letter—remains unclear (in 1985, Kitov recalled that his
letters to Party leadership were “jammed,” or strevali).19 We do know that
the second letter never arrived at the desk of Khrushchev, Brezhnev, or any
member of the first secretary’s inner circle like the first letter had. Instead,
the letter—with its criticism of the military and proposal to share military
technology with civilians—fell into the hands of his military supervisors,
who were infuriated. Without Berg’s support and under his military super-
visors’ initiative, a special military commission was convened to review
Kitov’s report.
The highly respected high-ranking war hero Field Marshall K. K. Rokoss-
ovsky chaired the commission as then chief inspector of the Ministry of
Defense. Rokossovsky—who survived the great purge, show trials, and
torture under Stalin—might have been sympathetic to Kitov’s case had he
actually attended the commission. As it happened, however, Rokossovsky
barely participated in the commission, leaving Kitov’s fate in the hands of
his supervisors, who rejected the proposal and followed standard Soviet
procedure in burning the unapproved (and irreproducible) proposal in
what colleagues later referred to as Kitov’s show trial.20 “Hence the paradox
in technics,” as Lewis Mumford put it: “war stimulates invention, but the
army resists it!”21
Incensed by Kitov’s critique, the unchecked commission exacted further
retribution by revoking his Communist Party membership for the follow-
ing year and dismissing him from military leadership, his position as the
director of Computational Center-1 of the Ministry of Defense, and effec-
tively his once meteoric military career. To justify this punishment, the spe-
cial commission deemed Kitov’s proposal “inefficient” for having suggested
that civilians should use of military technologies, disregarding any discus-
sion of his promised cost savings and efficiencies. The commission also
issued a formal complaint against Kitov for not having filed his network
proposal according to proper protocol. He was to be punished formally for
having attempted to send his communiqué to Khrushchev directly, bypass-
ing the intervening administrative tiers between him and the Party leaders.
Given the success of his first improperly filed letter to Khrushchev earlier
that year, a breach in filing protocol struck his colleagues as a disingenu-
ous and insufficient cover for severely punishing an army researcher who
was celebrated for having done something similar a year earlier. Eyewit-
nesses confirm that the commission’s unwritten response had little to do
with filing protocol, efficiency, or any other stated reason. Instead, they

9800.indb 88 6/2/16 3:05 PM

From Network to Patchwork  89

underscored the military’s possessiveness and unwillingness to share infor-

mation technology with the civilian sector. As Kitov later exclaimed in an
interview, the commission’s unwritten response was essentially that “the
army will never occupy itself with fulfilling any tasks concerned with the
national economy!”22
Most notable about Kitov’s show trial is not the possessive self-inter-
est that motivated a large institution to punish its own—an instinct that
animates most centralized command-and-control administrations, includ-
ing the military portion of the Soviet knowledge base—but rather the
counter-innovational institutional conditions that sanctioned the mili-
tary command to separate military and civilian resources, both economic
and technological. The military top brass decided to hoard its computing
resources, neither acknowledging the Politburo’s support of Kitov’s pro-
posals nor concerning itself with any commercial or civilian application.
Kitov’s military supervisors were free to act as they pleased, denouncing
any time sharing of their computer networks with others, even if doing so
at night would have had no obvious cost to the military and would work
against the interests of the top state leaders of the nation that the military
was sworn to protect.23
Kitov’s show trial also showcases the informal and contingent dynamics
that beset anyone trying to bridge the entrenched military-civilian divide.
Although Kitov’s first letter circumvented formal protocol without a hitch,
the second letter, which included military criticism, was intercepted. The
military man who was most likely to be sympathetic to Kitov’s case did not
attend the commission that he formally chaired. Informal degrees of free-
dom, in turn, allowed Kitov’s military supervisors, according to eyewitness
reports, to pronounce his proposed military-civilian nationwide network
an existential threat—not as much to the nation as to their personal and
unprecedented control over the resources of the nation.24 An automated
computer network threatened to automate and jeopardize the Ministry of
Defense’s positions of power over strategic bottlenecks of resources in infor-
mation technology, granting civilian economic planners access to the min-
istry’s technological monopoly.
Kitov’s first public computer network proposal ended his military career
and launched his career as a civilian network entrepreneur. In his many
publications promoting automated computer networks in the national
economy between 1959 and 1967, he continued to frame the economic
race in terms of military competition between the superpowers.25 In a 1959
article with Berg and Lyapunov, for example, Kitov announced to his read-
ers that the automation of firm-level economic management resulted in

9800.indb 89 6/2/16 3:05 PM

90  Chapter 3

major savings and “reductions in the administrative apparatus (in some

cases 80–90%).”26 In 1961, his advice to reform economics with computer
methods—without the military network—successfully secured the support
of the top Party leadership in the form of a Party report that helped pave
the way for Khrushchev at a Party Central Committee Plenum in Novem-
ber 1962. At that plenum, at the height of the cultural thaw and the eve
of the economic debates described previously, Khrushchev called for the
adoption of Western “rational” managerial techniques, proclaiming that
“in our time, the time of the atom, electronics, cybernetics, automation,
and assembly lines, what is needed is clarity, ideal coordination and orga-
nization of all links in the social system both in material production and
in spiritual life.”27 In many ways, influencing the leaders of the Soviet state
with cybernetic ambitions about networking the civilian economy proved
easier than bridging the military-civilian divide.
The Soviet military behaved as a well-oiled hierarchy when it limited
scientific or technological transfer outside of itself, although like the econ-
omy and Party apparatuses, its internal affairs could be unpredictable and
tenuous. Kitov’s case raises the point that so long as the military did not
have to associate its resources with nonmilitary projects, it was content
to manage its own internal affairs however it wished. So long as the Party
agreed (and often when it did not), it behaved as a private household unto
itself. This unpredictability could swing for or against military personnel.
In a well-ordered, top-down hierarchical military, research scientists are not
usually expected to be able to send letters directly to the heads of state or
to influence state policy with those letters. At the same time, a well-ordered
military probably would not permit middle-level administrators to dismiss
a star scientist from the army for proposing cost-saving procedures that
already were supported by the heads of the state, and if such a trial did take
place, the appointed dignitaries surely would attend and dismiss the case.
Yet none of this happened to Kitov, among untold others—and no one
found these events unusual.28

The Historical Concurrence of Cold War Networks

Because international communication networks precede national com-

puter networks, multiple network projects often emerge in very different
places at about the same time, and priorities are often the last thing to be
prioritized. By my accounting, Kitov in the fall of 1959 was the first to pro-
pose a national computer network for civilian communication anywhere,

9800.indb 90 6/2/16 3:05 PM

From Network to Patchwork  91

although this or any other “first” claim ignores the complex interdepen-
dencies of institutions and individuals that create any major technologi-
cal project. The rush to make “first” claims usually is seen in histories of
technological invention (especially histories written by retired professional
technologists) to enhance biographical hagiography and ignore claims
made elsewhere. It also can be difficult at the edge of any innovation to
distinguish between a slight improvement to an old technology and an
altogether new technological invention. Kitov’s EASU, like most of the pro-
posals examined here, assembled a network out of preexisting and new
telegraphy, telephone, radio, and radar networks. Rather than thinking of
them as computer networks, EASU was framed more as telephone networks
with computers. Simultaneously, the Soviet military, including computer
network designer Nikolai Matiukhin, knew of and sought to imitate the
automated air defense radar network that went operational in the United
States in 1958, although little about the classified SAGE project or its classi-
fied Soviet equivalent filtered into civilian science.29
Thoughts about ambitious civilian networks were percolating elsewhere
as well. Just months after Kitov’s second letter, the American psychologist
J.C.R. Licklider’s 1960 essay “Man-Computer Symbiosis” featured a vision
of the potential social and civilian benefits of computers, although (with
one footnoted exception) his essay restricts itself to local human-computer
intersections. In that footnote, he “envision[s], for a time 10 or 15 years
hence, a ‘thinking center’ that will incorporate the functions of present-day
libraries.” From here, “the picture readily enlarges itself into a network of
such centers, connected to one another by wide-band communication lines
and to individual users by leased-wire services. In such a system,” Licklider
concludes, “the speed of the computers would be balanced, and the cost of
the gigantic memories and the sophisticated programs would be divided
by the number of users.”30 In 1963, Licklider scaled up his vision of that
network as a library with an internal memo that was titled (half in jest)
“Memorandum for Members and Affiliates of the Intergalactic Computer
Network” and that sketched out the system that became the ARPANET—the
technical predecessor to the Internet.
Despite their historical concurrence, all available evidence signposts that
the early Soviet economic networks and the ARPANET developed indepen-
dently of one another. When the ARPANET went online in 1969, it took
the Soviet state by surprise. I have encountered no evidence to imply that
Kitov or others knew about Western computer network developments other
than the SAGE project. Nor have I found evidence that the American secret

9800.indb 91 6/2/16 3:05 PM

92  Chapter 3

intelligence community knew about Soviet cybernetic developments before

1964, when Soviet specialists at the CIA began wringing their hands about
Soviet cyberneticists working on a nonmilitary “unified information net.”31
Although subsequent Soviets would first pioneer socially ambitious nation-
wide network projects, the front of Soviet network projects, like the science
of cybernetics that underwrote it, proved anything but unified.
Other forms of international influence did lead to networks elsewhere.
In October 1957, for example, Soviet authorities set into motion events that
led to the ARPANET. Soviet rocket scientists used a missile to launch the first
manmade object into terrestrial orbit—Sputnik I, the first artificial satellite.
At the height of the cold war space race, Sputnik came after a number of wor-
rying developments. In November 1955, the Soviets air-dropped their first
thermonuclear nuclear bomb, during a time of tension when many Ameri-
can military strategists believed, probably incorrectly, that the Soviet fleet
of long-range bombers could reach American targets. The “bomber gap”
crisis in the mid-1950s, which was unfounded but drove defense spending,
launched that gap into orbital space. With Sputnik in orbit, the natural next
step was as obvious as it was terrifying: if a warhead were placed atop such
satellites, the world could be destroyed in a matter of minutes.
In February 1958, five months after the Sputnik crisis, the United States
Defense Department created the Advanced Research Projects Agency
(ARPA). This new government agency was charged with investing in and
advancing the frontiers of technology research beyond the immediate
needs of the military, especially in the spheres of space, ballistic missile
defense, and nuclear test detection. ARPA did not stay focused on milita-
rizing space for long, however. Two years after its creation, ARPA ceded its
space research jurisdiction to the distinctly civilian mission of the National
Aeronautics and Space Agency (NASA), which also was founded in 1958.
ARPA research then turned toward supporting basic, high-risk, and long-
term military research in information processing and computer systems for
tracking nuclear threats in the age of Sputnik.32
The focus on basic computer research questions made ARPA an opti-
mal site—under the auspices of the U.S. Department of Defense, a para-
gon example of a command-and-control hierarchy—for open-ended
basic research. Early computer innovations advanced by ARPA researchers
include distributed networking, time sharing, and packet-switching tech-
nologies (noted below). In 1965, shortly after President Lyndon B. Johnson
called for “creative centers of excellence” to advance basic research among
universities, the Department of Defense recommended using the ARPANET
to connect preexisting, government-supported computer research sites

9800.indb 92 6/2/16 3:05 PM

From Network to Patchwork  93

across the American academy—first at the University of California at Los

Angeles, Stanford University, UC Santa Barbara, and the University of Utah
and then eastward to the Massachusetts Institute of Technology, Carnegie
Mellon University, Harvard University, and other universities.33 The Soviet
military-civilian divide barred similar wide-scale collaboration between
defense projects and university contractors.
The ARPANET went online on October 29, 1969, as the first large-scale,
dual military-civilian use, packet-switching computer network in the
world, the “Mother of all nets” as it has since been known. In its first stage,
the ARPANET consisted of leased telephone lines and modems connecting
computer terminals at UCLA, Stanford, UC Santa Barbara, and the Uni-
versity of Utah. The first message sent was the prophetic utterance L and
O—“lo,” not as in “lo and behold” but as in the first two letters of the word
login that could be sent before the network crashed.34 ARPA directors in the
1960s negotiated careful balances between Congress (to whom the directors
promised research that could be applied to national security issues) and aca-
demic research contractors (to whom the directors promised the freedom of
basic research that would be independent of any defense rationale).35
The heyday of military research in the 1960s came to an end in the
political wake of the Vietnam War when in 1969 the first Mansfield amend-
ment curtailed military spending on science across the board and in 1973
the second Mansfield amendment dramatically limited ARPA funding to
appropriations for research directly related to military applications. ARPA,
stripped of the capacity to do basic research, saw many researchers migrate
to a fledgling computer industry, most famously Xerox PARC. Such a brain
drain or labor migration from the military to the civilian sector would have
had to be directed by military and state oversight in the Soviet Union.
So although in both superpowers the early computing industries in the
1950s through early 1970s depended on military state projects (with pri-
vate contractors used as spinoffs as well as employed by the U.S. Air Force
as its research consultancy), the biggest advantage that the United States
wielded over the USSR appears to have less to do with the market indepen-
dence of the private commerce than the porousness of research, resources,
and knowledge flows between military and civilian projects. The modest
and mixed military-civilian origins of the ARPANET are worth bearing in
mind as well: the ARPANET was designed and launched explicitly for civil-
ian scientists to exchange data at a distance. Its affordances as a network
for national communication became obvious after the fact with the inven-
tion of email in 1971. At the same time, these civilian, public networked
computing utility services were initially funded because of the military

9800.indb 93 6/2/16 3:05 PM

94  Chapter 3

justifications to design, fund, and build a nationwide communication net-

work that could survive a nuclear attack by the Soviets. This military moti-
vation led Paul Baran’s innovations at RAND in distributed networking and
packet-switched networking and distinguished the ARPANET from other
networks of its time. The emerging thesis here appears to be that the virtue
of the military-industrial-academic complex in the United States rested on
not the state, the market, or civilian research but on the complex that con-
nected these sectors.36
Meanwhile, Chile under Salvador Allende (1970–1973) and France in
the 1980s developed large-scale national networks. Unlike the strict Soviet
divide between military and civilian research and more like the far more
synthesized “military-industrial-academic complex” in America (perhaps
the most important element of that phrase for understanding midcentury
big science are the hyphens), the cases of Chile and France show that the
international history of civilian networks cannot be easily separated from
that of military networks. The Soviet military tightly siloed its technical
innovations, the East German Stasi shuttered its large-scale computer net-
work capacities from serving and transferring to civilian applications, and
the West German government also forbade the transfer of network capaci-
ties from military to civilian.37
No country escaped institutional frustrations in developing nationwide
computer networks. At important times, the complex in cold war American
science proved vexatious, if not impossible, to navigate. Take, for example,
Paul Baran (1926–2011), a Polish-born engineer who was raised in Phil-
adelphia and Boston. Baran is widely remembered today for innovating
packet-switching and distributed-network designs, which now are central
to modern-day networking, but his struggles are less well remembered. In
1960 at the RAND Corporation, a research think tank under contract with
the U.S. Air Force, Baran articulated the “hot-potato heuristic” behind mod-
ern-day data traffic on the Internet: break down a message into packets (or
envelopes) of information, release each packet to travel on its own traffic-
reducing pathway to its final destination, and resequence and receive all
packets in their original order. In the early 1960s, Baran also designed the
celebrated idea of a distributed network in which every node in a network
connects to its neighboring nodes and not to any decentralized or central-
ized node arrangement (figure 3.2).
Widely celebrated as a prototype to “end-to-end” intelligence and a
liberal democratic mode of communication, Baran’s network innovations
were colored and shaped by the cold war military complex as well as cyber-
netic sources. In the embarrassing aftermath of Sputnik, the U.S. Defense

9800.indb 94 6/2/16 3:05 PM

From Network to Patchwork  95



(a) (b) (c)

Figure 3.2
Three network types: (a) Centralized, (b) decentralized, and (c) distributed. Source:
From Paul Baran, “Introduction to Distributed Communication Networks.” On Dis-
tributed Communications, RAND Corporation Memorandum RM-3420-PR, August
1964, 2. Reproduced with permission of The Rand Corp.

Department ordered ARPA to design a “survivable” network that would last

long enough in a nuclear strike to send a “go-code” to guarantee “second-
strike capability.” “There was a clear but not formally stated understand-
ing,” noted Baran, “that a survivable communications network is needed to
stop, as well as to help avoid, a war.”38 A network that can survive an enemy
attack could ensure the threat of the mutual nuclear annihilation—a threat
so cataclysmic that it would rationally deter (Baran and his military supe-
riors hoped) either the Soviets, the Americans, or any other nuclear power
from striking first.39
Baran’s inspiration for packet switching as a way to build a survivable
network traces back to Warren McCulloch’s cybernetic conception of the
human brain as a complex and resilient logical processor. As Baran reported
in an interview with Stewart Brand, “McCulloch in particular inspired me.
He described how he could excise a part of the brain, and the function
in that part would move over to another part.”40 The same interview lists
McCulloch and Pitt’s 1943 paper on neural networks as a sensible refer-
ence, although Baran also noted that he was reading more broadly in the
“subject of neural nets,” a literature that probably included McCulloch,

9800.indb 95 6/2/16 3:05 PM

96  Chapter 3

Pitts, Jerome Lettvin, Humberto Maturana, and others. Much of this char-
acterized “McCulloch’s version of the brain,” which, Baran continued, “had
the characteristics I felt would be important in designing a really reliable
communication system.”41 Reliable national computer networking were
inspired by models of complex (heterarchical) neural networks.
The result of Baran’s conversations was packet switching, a technology
that broke messages into “packets,” which allowed digital “bursts” of data
to be rerouted around damaged parts of a network—just as the brain can
reroute neural impulses around damaged neural matter. Similarly, Baran’s
observation was that, due to network effects, the brilliance of a distributed
network, whether neural or national, is that it does not need each of the
average eighty-six billion neurons in the human brain to connect to every
other (and the number of possible connections between eighty-six billion
neurons is so incomprehensibly large that the need for robust reconnection
becomes obvious).42 Rather, attaching to a couple of other nodes allows a
distributed packet-switched network to reroute in real time around dam-
aged territory, whether neural or national.
The governing logic behind Baran’s innovations is curiously the same as
McCulloch’s heterarchy: in a heterarchy, the relations between nodes can
be ordered and evaluated in more ways than one, and there is no overarch-
ing governing structure, no internal logic, and no accounting regime for
determining how nodes interconnect. Both lack a fixed control center or
mother node. Baran did not concern himself with theorizing about a dis-
tributed communication network as a neural network for the nation, as a
cyberneticist might. McCulloch’s ideas about the brain as a self-governing
network helped Baran to arrive at concrete pragmatic solutions to the over-
arching military orders of his employer. The Internet, in this sense at least,
traces its intellectual sources back to cold war cybernetics.
Baran’s network innovations do not arrive without serious institutional
and international complication. Although technically on target, Baran’s
ideas were not influential until after a foreigner—an Englishman named
Donald Davies, with the UK Post backing him—independently discovered
and articulated packet switching. Only then did Baran’s superiors in the
U.S. military-industrial complex start paying attention to his ideas. In fact,
between 1960 and 1966, AT&T repeatedly declined or delayed his propos-
als to develop digital communication networks. As one AT&T official told
him, the near nationwide monopoly on analog telephony networks was
not about to go into competition with itself. When it appeared that the air
force stood ready to implement Baran’s ideas without AT&T, Baran with-
drew his proposal because he felt that the appointed government agency,

9800.indb 96 6/2/16 3:05 PM

From Network to Patchwork  97

the Defense Communication Agency, would “screw it up and then no one

else would be allowed to try, given the failed attempt on the books.” With
no such “competent organization” in sight and after spending six years
aggressively publishing his network research internationally to ensure max-
imum circulation about how survivable communication networks could
help ensure mutual deterrence, Baran despaired at the local prospects and
turned his attention elsewhere.43 The popularity of the phrase packet switch-
ing, which was Davies’s term, and the obscurity of Baran’s initial coinage
block switching are evidence that it took outside competition to spur local
authorities to take packet switching seriously. The U.S. ARPANET, despite
the efforts of its own network entrepreneurs, was inspired by foreign found-
ers. To the degree that Stigler’s law of eponymy holds—“no scientific dis-
covery is named after its original discoverer” (a law that Stigler attributes
with a grin to Robert Merton)—Baran’s case rehearses not the exception
but the rule that international communication networks precede national
computer networks.

Aleksandr Kharkevich’s Unified Communication System (ESS)

At the same time that Paul Baran was publishing his network research in
the hopes of ensuring the Soviets would have access to survivable commu-
nication networks and that J.C.R. Licklider was thinking about computer
networks as pragmatic tools for facilitating long-distance exchange of sci-
entific data, Soviet cybernetic network entrepreneurs were imagining com-
puter networks as ambitious infrastructural solutions to the nation’s most
pressing civilian problems. In the imaginative minds of the three Soviet
cyberneticists chronicled below, digital computer networks were models
both of and for the entire nation.44 These and other early proposals for a
“unified system of calculating centers for the development of economic
information” found their earliest inspiration in the 1955 Academy of Sci-
ences proposal by Vasily Nemchinov (two years before Sputnik and well
before the invention of the ARPANET) that considered erecting large but
unconnected state computer centers (in Moscow, Kiev, Novosibirsk, Riga,
Kharkov, and other major cities) that could facilitate the local exchange of
scientific reports and economic information among regional economists.
One of those proposals—Kharkevich’s unified all-state system for informa-
tion transmission—has been relatively neglected in previous commentary
and receives additional attention below.
In 1962, Aleksandr A. Kharkevich, then deputy chair of the Council on
Cybernetics, proposed a communications network for the entire nation,

9800.indb 97 6/2/16 3:05 PM

98  Chapter 3

although this proposal—for a network that was formally called the “unified
all-government system for the transmission of information” (edinaya obsh-
chegosudarstvennaya sistema peredachi informatsii)—did not seek to solve an
explicit civilian-sector problem, unlike other contemporary Soviet network
projects. In fact it sought to solve no particular problem at all: it was pro-
posed out of sheer technical ambition to build a national communication
network on preexisting telephony and telegraphic channels for all kinds of
data exchange. The closest that Kharkevich comes to a social justification is
noting, without comment, that it will “broaden the sphere of human activ-
ity.”45 The technical orientation of data exchange in Kharkevich’s proposal
resembles the purpose of the ARPANET as a network for exchanging data
between scientists. With similar “intergalactic” ambitions, Kharkevich set
out to optimize all technical communication problems at once by propos-
ing to merge all Soviet data streams into a single nationwide digital com-
munication network. His 1962 proposal came to light in an article titled
“Information and Technology” that was published in the leading periodical
Communist, in which Kharkevich apparently renamed this network with
the more workable title of “unified communication system” (ESS, for edi-
naya sistema svyazi), a possible source of the uncited CIA speculations about
a menacing Soviet “unified information net.”46 His vision describes a tech-
nical future that was obvious to information theorists, who were the tech-
nocratic twin of cyberneticists and could be traced back to Claude Shannon
of Bell Labs and his seminal 1948 article “A Mathematical Theory of Com-
munication.” (Kharkevich was himself a leading information theorist and
specialist in noise reduction in electronic communication signals.)47
In the 1962 Communist article, Kharkevich proposes that the ESS uni-
fied network of information transmission be built, like the other proposals
here, on the preexisting telephone and electronic network infrastructure,
which he found analogous to a nationwide railway network that was built
to transmit, store, and process digital information messages. Given that
most Soviet citizens had to use public phones in 1962, this was a fantasti-
cally far-fetched technical proposal on any terms. Perhaps for this reason,
he devotes almost the entire twelve-page article to the technical capacities
of such a network (for example, how messages would arrive at the right
place without data loss), which were grandiose. Telegraph cables, telephone
lines, radio waves, and all other technical communication channels were
to be unified into a common digital and “enciphered” currency that would
be related by binary electronic pulses over telephone wires. The irreduc-
ible denominator to his technocratic vision was the concept of informa-
tion: “the far-reaching role of information has become clear not only in

9800.indb 98 6/2/16 3:05 PM

From Network to Patchwork  99

the relations between people, but also in the interactions between man
and machine, as well as in the life of any organism.” He continued, “with
the enhancement of economic, technical, and cultural levels of society, the
amount of information necessary to collect, transmit, and somehow pro-
vide for all functions of the community of people grows faster and faster.
No organized form of activity is thinkable without information exchange.
Without information, planning and governance are impossible.”
Backlit by the stated universal need for information, Kharkevich justified
the network proposal by citing the “prominent system ‘SAGE’” computer
system in the United States and Canada as a parallel to his vision of a nar-
rowly applied, universal information system for antiaircraft defense. The
top of his pyramid, ESS network design, was meant to “fulfill the func-
tion of the dispatcher of the network,” or “the center will be constituted
by a large group of specialized calculating-logic machines, appointed for
the direct resolution of the many changing conditions of one single task:
the supply of increasingly favorable conditions for the appointment of all
currency flows of information.”48 The Soviet Union did need not to stop
at antiaircraft defense, he said, concluding his “grandiose thought” of an
all-reaching, full-service ESS network with the observation that “creating
an all-state unified system of connection … would only be possible in a
socialistic government under the conditions of a planned economy and
centralized government.”49
Kharkevich’s article is remembered among some technologists today not
for proposing the ESS but for formulating what became known as Kharkev-
ich’s law. This law holds that the quantity of information in a country grows
proportionally to the square of the industrial potential of the country (N2).
The original formulation of his observation in the article is perhaps less
elegant than information technologists might remember: “Given a large
number of factories, the number of paired links between them is approxi-
mately equal to half of the square of the number of factories.”50 The law,
in effect, prophesies a power law connection at the macro level between
an industrial society and an information society. In 1965, the American
computer businessman Gordon Moore expressed a distinct exponential law
that has applied to the microscopic level of the compounding growth of
silicon chip production—that the number of transistors on an integrated
circuit doubles every two years (2N).51 Both men foresaw in 1962 the emerg-
ing information sector or what Austrian American economist Fritz Machlup
called “the knowledge economy.” For Kharkevich, the amount of informa-
tion that a society processes can be expressed as a power law function of
the industries it contains, and for Moore, the amount of information that a

9800.indb 99 6/2/16 3:05 PM

100  Chapter 3

society processes can be expressed as an exponential function of the transis-

tors on the circuits its industries can produce.52 These sibling laws (Moore’s
2N and Kharkevich’s N2) diverge interestingly in complex systems (when N
is larger than 4). They also backlight their micro and macro focuses—Moore
on microscopic industrial production and Kharkevich on informational
industrial society. The result is different framings of the national network
as a sort of central processor. Like Baran, Kharkevich prioritizes building
“survivable” military networks while also looking to benefit other civilian
and social goals.
Unlike the cybernetic metaphor of the brain that Baran drew from War-
ren McCulloch, Kharkevich saw his national network as a nervous system
that was overlaid onto the body of the nation and that would be governed
by a central processor, or brain, located in Moscow. The resulting contrast
of cybernetic metaphors for the information societies is again sharp: for
Kharkevich, the networked nation was the body controlled by a central
brain, and for Baran, the networked nation was the brain itself.
Like other network designers, Kharkevich also designed the ESS network
after the formal administrative structure of the nation that he imagined it
would network. “It is natural that the network should be supervised,” he
wrote, “by the Ministry of Communication [Svyaz’] in the Soviet Union,”
the ministry that managed many preexisting networks for information
exchange, including telegraph, telephone, phototelegraph, messages
(courier), and early digital technologies then available in small numbers.
Kharkevich breezily dismissed the distributed network model that Paul
Baran was developing at the time (although not by that name), observ-
ing that the structure of the network needs to be able to connect any two
nodes, and he writes, “in order to do this, it goes without saying, one does
not need to unite all nodes with separate lines.”53 Instead, Kharkevich con-
sidered a hierarchically decentralized design, or pyramid structure, “the
rational structure of a network.”54 Like transport roads, his network would
split out in a “radial system” in which a “given territorial group is united by
links to a communication node.” Just as every local, regional, and territorial
group would have its own common node, Kharkevich was quick to stress
the center that was implicit in this “radial” design. In 1962 in Moscow, a
“centralized automated management” design would have appeared reason-
able, if still monumental in aspiration, to him:

The brisk carrying out of these functions is possible only … if the entire network will
work under centralized automated management. The governing center of ESS should
distribute information about the state of the network at every given moment.…
The center should be capable of predicting such changes [in the network traffic]

9800.indb 100 6/2/16 3:05 PM

From Network to Patchwork  101

slightly in advance and create the needed operating reserves. The center appoints
the pathways for the passage of flows of information as they depend on the general
state of the network; in the case of necessity [nadobnost’], the center will be able to
focus all network resources on fulfilling special information transmission tasks. In
short, the governing center of ESS fulfills the function of the dispatching manager
of the network.55

Here, without the benefit of packet-switching protocols, Kharkevich

anticipated the needs of a nationwide network to adapt automatically in
real time to traffic jams as well as the capacity to complete “special informa-
tion transmission tasks” (the sending of nuclear “go-codes” in the case of
nuclear “necessity”). Automation appears to be the ultimate nuclear safe-
guard, for he continues that “it will not be possible to give these functions
to people. The center will constitute a large group of specialized calculating-
logical machines, appointed for the direct resolution of the changing con-
ditions of a single task: providing increasingly favorable conditions for the
appointment of current flows of information.”56
The fate of the ESS owes less to the technicalities of its design than to
the muses of institutional historical contingency. In 1961, the year before
he proposed the ESS, Kharkevich was made director of the Institute for the
Problems of the Transmission of Information (IPPI), the new Soviet Acad-
emy of Science’s research center on information technology. The president
of the Academy of Sciences, Mstislav Keldysh, a rocket scientist and math-
ematician who helped develop the Calculation Bureau during World War
II, created Kharkevich’s IPPI in the same year that he created Glushkov’s
Institute of Cybernetics in Kiev and Fedorenko’s Central Economic-Mathe-
matical Institute in Moscow. In 1963, Kharkevich’s ESS vision took its first
step forward when the Ministry of Communication created an interagency
Coordinating Council, chaired by the then minister of communications,
General-Colonel N. D. Psurtseva, to supervise the creation and standard-
ization protocols for the ESS. However, before any the council could make
concrete progress and three years after proposing the ESS, the project col-
lapsed. On March 30, 1965, Kharkevich died of protracted health problems
at the age of sixty-one. Why no one took up his reigns on the ESS proposal
remains unclear, although the lack of evidence implies that the ESS’s politi-
cal prospects passed into history with Kharkevich.

N. I. Kovalev’s Rational System for Economic Control

Consider still another short-lived and concurrent network proposal,

whose fate archival materials and interviews have not yet clarified. In the

9800.indb 101 6/2/16 3:05 PM

102  Chapter 3

November 1962 Plenary Meeting of the Communist Central Committee,

decisions were made to mechanize and automate both the industrial pro-
cesses and the administrative control over those processes. In the 1963
issue of Problems of Economic Transition, N. I. Kovalev, then the director of
the State Economic Council (Goseconomsovet), published a proposal that
elaborated on those decisions and proposed creating and connecting the
preexisting major computing centers for each of the regional economic
councils (sovnarkhozy) that Khrushchev initiated in 1957. Like all the oth-
ers, Kovalev’s design also mapped a pyramid communication network onto
the economy’s three-tier hierarchy of ministry, regional council, and local
enterprise. The network was meant to help the regional councils to receive
otherwise unspecified “necessary information” on time. No longer would
“the report materials arrive so late that they cannot be effectively used to
plan and govern the national economy.”57 Citing Nemchinov and Glush-
kov (prominent specialists in the field who are featured in the next chap-
ter), Kovalev estimated that the network would cost 94 million rubles, the
first layer of thirty computing center would require three years to complete,
and the economic savings would far outweighing the costs.58 By referring to
such a computer network as a “rational system,” Kovalev did not emphasize
the transformative effects of long-distance real-time computer networks
but instead stated a need for vaguely specified “cybernetics, electronic com-
puting and control devices” to serve as the “material and technical base”
for a transition to a communist model for “planning and controlling the
economy” over the next two decades.
Kovalev’s proposal stands as a synecdoche for a larger competition
among the cybernetic and mathematical economists on one side and state
planning agencies and party leaders on the other. Both economic plan-
ners and party leaders advanced arguments for and against the comput-
erization and networking of the command economy in terms of whether
technocratic reform would lead to the proper control over information.
Kovalev, together with his cybernetic colleagues and allies, saw in the net-
worked computer a grand manipulator for transforming the economy as a
giant information system in need of optimization, objective planning, and
diminishing bureaucratic overhead costs. Curiously, the most influential
opposition to such proposals came from the main planning state agencies,
including Gosplan, the Central Statistical Administration (CSA), Gossnab,
and regional and branch committees. These groups openly resisted his and
similar proposals because they were perceived to involve personal loss of
control over the information in the command economy.

9800.indb 102 6/2/16 3:05 PM

From Network to Patchwork  103


By 1963, with three national network proposals already on the table—

Kitov’s EASU, Kharkevich’s ESS, and Kovalev’s unabbreviated “rational
system of economic control”—the institutional landscape was evolving
toward some kind of head in economic reform. That intellectual terrain
includes various supplemental network projects that promoted the core
Soviet cybernetic instinct that large-scale information systems, such as the
command economy, can become self-sustaining and even self-governing
systems. It may be helpful to distinguish between two meanings of the
word automated—(1) having operations that are entirely independent of
human involvement and (2) having operations that are designed to receive
and interact with humans but do not necessarily need human involve-
ment. The OGAS, understood as an explicitly cybernetic human-computer
interface, clearly signals the latter sense of the term.59 In other words, the
conceit of cybernetic (human-machine) self-sufficiency was not to imag-
ine a national economy that was independent of any other outside forces
but rather to envision a socialist planning apparatus that engaged with the
economic body it networked and that, together, would prove responsive,
balanced, and self-governing. By contrast, the liberal economists sought
a different path to self-governing markets—introducing profit measures
into local enterprise accounting while still maintaining basic production
guidelines for the overall economy. Both cybernetic and economic liberal
reforms reached compromise solutions with the operations of the com-
mand economy, just in opposite directions. The cybernetic economists
offered a technocratic reform that was meant to work with human admin-
istrators and liberal economists—a market reform that was meant to work
with command economy guidelines.
These contending approaches came to a head in 1963 through 1965 at
the same time as the bumpy transition of state power from Khrushchev to
Brezhnev. Because both approaches to reform met with unsystematic but
widespread resistance from orthodox economic planners and professionals
who were comfortable in their current positions, both produced tentative
heirs to the economic debates in the early-mid 1960s—the OGAS proposal
in 1963 and the Kosygin-Liberman reforms of 1965. Early Soviet networked
computing culture was decentralized in practice, despite the state’s central-
ized design in principle. Kitov’s EASU first proposed having technomili-
tary networks be put to public and social benefit, but he found himself
grounded for attempting to bridge the yawning military-civilian divide.
Three years later, Kharkevich’s ESS, with Kovalev following suit, reached

9800.indb 103 6/2/16 3:05 PM

104  Chapter 3

ever further, networking all technical signals into a network resembling the
pyramidal state while staying silent about any social ambitions. Yet these
early proposals fell prey to strategic veto points in the state administra-
tion that depended not on bureaucratic rules but on charismatic leader-
ship and personal power. The technical open-endedness of Kharkevich’s
ESS probably most closely resembles that of the ARPANET, although the
ARPANET began with the modest goal of scientific data exchange and the
ESS, like the others, began with an ambitious blueprint for an entire digital
nation. Unsurprisingly, the more ideologically charged economic networks
faced more ideological opposition, and the fate of the ESS points to the
charismatic actor-dependent institutional disorder that governed the Soviet
knowledge base.
Perhaps the signal lesson to take from these early Soviet network propos-
als is that there is no inherent connection between the designs of techno-
logical and political systems. Many digital theorists in liberal democracies
have imagined the effects of technology in the terms of their local political
systems, claiming that digital technologies must be deliberative, direct, and
participatory—similar to that of contemporary democracy discourse. So,
too, did these Soviet cybernetic theorists imagine that a nationwide com-
puter network would “naturally” map onto the design biases and design
logics of the formally top-down centralized administrative hierarchy of the
Soviet state. Both visions are theoretically imaginative because they neglect
actual political practices and their significant costs and consequences.
These network proposals ignored the informal, nonhierarchical functions
of the Soviet state and society, just as modern democracies involve far more
than just the representation of individual voices celebrated by many digital
media theorists. Centralizing computer networks and centralized socialist
states have as little to do with one another as the digital does with democ-
racy. Both propose imaginatively rich associations about what could be,
promising no less than some pseudo-automatic or pseudo-democratic form
of self-determination, but do little to affect careful or accurate assessments
of how politics actually works on the ground.60
These waves of cybernetic imagination about the fit between computer
network and formal state and social structure repeatedly broke against the
rocks of widespread practice that countered the official Soviet imagination
of itself. Paul Baran struggled to secure institutional support from American
corporations and a state that refused to recognize the value of what became
the key network innovations of his age. So, too, in Moscow, Kitov had to
abort the EASU due to his unsuccessful attempt to bridge the abyss that
separated military and civilian research, Kharkevich’s ESS collapsed with

9800.indb 104 6/2/16 3:05 PM

From Network to Patchwork  105

the health of the man appointed to steer its grandiose technical ambitions,
and Kovalev’s “rational system” fell short of convincing his colleagues in
economic administration that ceding their own decision-making power to
automated computers could either rationalize or systematize the work of
economic planning. All four of these early network projects—three in the
Soviet Union and one in the United States—did not take shape due to an
imagined and often misleading connection between political and techni-
cal systems. All four rooted their imaginations in the explicitly cybernetic
terms of analogizing across technological and social systems. This imagina-
tive and at times utopian instinct for political-technological system analogs
leads theorists to neglect the significant costs and consequences that come
from actual political practice. As it often happens, the revolutionary reach
of our modern technological imagination of large-scale networks (among
other things) often ends up serving local institutional self-interests and the
status quo. The next chapter extends and complicates this theme in its his-
tory and analysis of the central and longest-lasting attempt to network the
Soviet Union.

9800.indb 105 6/2/16 3:05 PM

9800.indb 106 6/2/16 3:05 PM
4  Staging the OGAS, 1962 to 1969


The year 1962 proved to be a tumultuous one for the world. Khrushchev’s
grasp on the reigns of the Soviet state began to slip in the face of mounting
criticism, and Kennedy’s Bay of Pigs invasion metastasized into the Cuban
missile crisis, probably the closest the world has yet come to a nuclear world
war.1 Behind the scenes to these potentially cataclysmic situations, a small
team of Soviet cyberneticists who were located in Kiev and Moscow were
committed to building “electronic socialism” under the guise of the All-
State Automated System, or OGAS. The OGAS Project was the Soviet Union’s
attempt to build a national computer network project that would network
the command economy, automate and optimize the immense coordination
problems besetting that economy, and thereby speed the grand socialist
experiment toward a prosperous and stable Communist future.
The All-State Automated System Project took its first breath with the
delivery of a sealed envelope into the hand of Nikita Khrushchev in the late
fall of 1962. The letter to the general secretary was written by young scien-
tists from the Komsomol Spotlight (Komsomol’skii prozhektor), who noted
what they perceived to be the catastrophic backwardness of information
technology in the USSR compared to the United States and called for the
immediate acceleration and adoption of computing technology into eco-
nomic planning. The letter made an impression on the public, in the form
of an official Izvestiya newspaper article titled “Information Technology in
the National Economy,” and on members of the Politburo, the governing
committee of the Soviet state, which reportedly spent nearly thirty-five
minutes of a forty-five-minute session discussing the consequences of their
fifteen-page letter. Several months later, on May 21, 1963, following the
proposals discussed in the previous chapter, the Politburo with the backing
of all relevant ministers advanced a Communist Party resolution calling for
the same and authorizing the first economic reform carried out by auto-
mated computer network (later known as the OGAS). This chapter discusses

9800.indb 107 6/2/16 3:05 PM

108  Chapter 4

the vision, the chief visionary (Viktor Glushkov) and his team, and the
institutional landscape for the OGAS Project, the most prominent attempt
to establish a civilian national network project in the Soviet Union.

The OGAS: A Vast Vision behind a Global-Local Network

The OGAS Project promised to deliver “electronic socialism” that was as

ambitious as its official title was long—the All-State Automated System for
the Gathering and Processing of Information for the Accounting, Planning,
and Governance of the National Economy, USSR. Its short names were the
All-State Automated System for the Management of the Economy, the All-
State Automated System, and OGAS. For clarity, I distinguish here between
the OGAS as the imagined network that did not come to exist and the OGAS
Project as the Soviet actors and institutions that tried to realize this reform.
According to its cyberneticist founders, the infrastructure of the com-
mand economy had to be upgraded before the entrenched coordination
problems that led to the country’s economic woes could be resolved. “In
the area of economic management,” Glushkov wrote in 1962, “cybernetics
fits our socialist planned economy like a glove.”2 The work was fundamen-
tally technocratic and rational and sought to “reduce the influence of the
subjective factor in the making of administrative decisions.”2
In its most modest framing, the OGAS—which stretched nationwide
across preexisting and new telephony wires that were entirely separate
from preexisting military computer networks—appears little more than the
extension of a local factory control computer network. It would be an ASU
(automated system of management) or OGASU (All-State ASU) (Obshche-
Gosudarstvennaya Avtomatizirovannya Sistema Upravleniya). The primary
visionary of the OGAS, Viktor Glushkov (who is discussed later in this chap-
ter), had been aware of Anatoly Kitov’s efforts, including his Red Book letter,
ever since Glushkov began studying computing in Kiev with Kitov’s 1956
Digital Computing Machines in hand. Glushkov employed Kitov as a consul-
tant in 1960 after Kitov’s dismissal from the army. The OGAS was to become
the Soviet equivalent of the national economy imagined as a single factory,
with one interactive industrial control system serving it across a national
computer network in real time. This was not to be a dumb network that
would merely exchange data and communication across great distances.
It was to be a “smart” network whose decentralized command and con-
trol protocols would be capable of automating, mathematically modeling,
optimizing, and rationalizing away the profound inefficiencies that beset
the command economy. According to the original proposers, the resulting

9800.indb 108 6/2/16 3:05 PM

Staging the OGAS, 1962 to 1969  109

network efficiencies, which were optimized to serve both national and local
needs, would achieve full effect by 1990, nearly thirty years after the Komso-
mol young scientists delivered their letter into Khrushchev’s hands.
As originally envisioned, the OGAS had several distinct features. Perhaps
the most meaningful contrast with that of modern networks is that the
OGAS was modeled after the economy of a factory writ large for a nation.
The basic unit of the OGASU was, as the initials imply, the ASU, the auto-
mated management system, or a local information and control system that
looped onsite mainframe computers into the industrial processes of a fac-
tory or enterprise to provide real-time information feedback, control, and
efficiencies. This kernel vision of a network as an expression of the nervous
system of a factory, writ large across a nation, magnified the image of the
workplace until it incorporated the whole command economy—a sort of
simultaneously metaphorical and mechanical collectivization of the indus-
trial household (or what Hannah Arendt calls the oikos).
The OGAS Project might be seen as preceding, although not precipitat-
ing, the current trends in so-called cloud computing. The national network
was to provide “collective access,” “remote access,” and “distance access”
on a massive scale to civilian users who could “access,” “input,” “receive,”
and “process” data related to the command economy (such older terms
appear to bear more descriptive heft than the modern computing meta-
phors such as upload, download, share, and stream). The decentralized net-
work was designed so that information for economic planning could be
transmitted, modified, and managed in relative real time up, down, and
laterally across the networked administrative pyramid. At the base of that
pyramid, in the network’s initial vision, were as many as twenty thousand
computer access points and ASUs distributed throughout the nation’s enter-
prises and factories. This base of computer centers would be connected to
one hundred to two hundred midlevel regional planning decision centers
in major cities, which would be connected to the central planning process-
ing center in Moscow by high-capacity data channels. The original vision of
a three-tiered pyramid network—with twenty thousand computer centers
on the bottom, one hundred to two hundred in the middle, and one on
the top—was scaled back in the original design of the technical base of that
network (the Unified State Network of Computing Centers, or EGSVT). The
first proposal for that technical network offered a modest blueprint where
one central computing center in Moscow would regulate only twenty-five
to thirty computing centers in city sites of “information flow concentra-
tions” and an unspecified number of “regional calculating center and
points of information gathering”3 (figures 4.1 and 4.2).

9800.indb 109 6/2/16 3:05 PM

110  Chapter 4

9800.indb 110 6/2/16 3:05 PM

Staging the OGAS, 1962 to 1969  111

Figure 4.1
Map of the three tiers (I, II, III) of planned computing center sites behind the OGAS
(All-State Automated System), 1964.

9800.indb 111 6/2/16 3:05 PM

112  Chapter 4

Figure 4.2
Map of the EGSVTs (Unified State Network of Computing Centers) that were pro-
jected to be operational in 1990, possibly from 1964.4

As communication scholar Vincent Mosco has recently noted, the Sovi-

ets offer perhaps the first glimpse of the modern imagining of decentralized
remote computing (what recently has been called cloud computing) on a
massive scale.5 In Glushkov’s design, the network would afford interactive
and collective remote access and communication vertically up and down
the planning pyramid and horizontally among peer and associated com-
puting centers. Glushkov writes: “the characteristic quality of the network
was a distributed database with zero-address access from any point of the
system to all the information after automatic verification of the qualified
user.” In other words, any user with proper permission could access all
the content of the network at any point on the network. At local levels,
factory workers would be able to input their own information, reports,
and recommendations about improving factory workflow, which would
automatically be stored in a national unified database for local, regional,
and national review. The content format was not to be prespecified. For
example, the network visionaries planned to include over 500,000 project

9800.indb 112 6/2/16 3:05 PM

Staging the OGAS, 1962 to 1969  113

dossiers on foreign scientists, engineers, executives, and companies in the

OGAS nationally networked database. (From 1963 to 1968, the associated
Department of Scientific Institutions gathered about 75,000 such dossiers.)
The proposal’s other ambitions went far beyond that of simply sluicing eco-
nomic planning information. In 1971, the deputy editor of Pravda, Viktor
Afanasyev, for example, reasoned that OGAS “can be used—and should be
used—for gathering, processing, and analyzing information on sociopoliti-
cal and ideological processes as well, for the purpose of optimal manage-
ment [of society].”6
As a near synthesis of optimal management and total surveillance, the
OGAS is a full articulation of the wider political-economic imagination
of the Soviet Union as not just a single unified society and set of nations
but as a unified corporation with a socialist mission statement. The OGAS
appeared to its founders as the information technology upgrade that the
Soviet Union had long needed to be able to function as the corporation it
had already long imagined its command economy to be—a single and com-
plex organization that featured decentralized means of control and com-
munication for circulating the informatics lifeblood of a socialist economy.
Because socialism openly recognized economic activity as more than merely
computational, the network that would best facilitate its fitness would also
control and communicate associated political and social concerns as well.
The OGAS Project of course was no ARPANET. It sought much more than
data transfer and communication among scientists. From the outset, the
OGAS Project sought to bring the economic bureaucracy online by mak-
ing all relevant government documents electronic, allowing a decentralized
remote access to all economic workers, and allowing decentralized access
for controlling and optimizing the information in those documents. The
decentralized design of the network project is worth stressing. Although
still hierarchical, acquiescent to Moscow as the center, and state-led, the
longest-lasting Soviet network proposal was (unlike the full central control
in Kitov’s EASU and the radial design of Kharkevich’s ESS) openly worker-
oriented, antibureaucratic, and decentralizing in principle. This gives the
OGAS Project and its team more credit than many commentators and critics
have given it. Both international and internal critics, including the British
organizational cyberneticist Stafford Beer, were critical of Soviet manage-
ment techniques.7 More than a network, the OGAS Project as formulated
by Glushkov outlines a daring technocratic economic imagining that was
meant to operate in a future Soviet information society by digitizing, super-
vising, and optimizing the coordination challenges besetting the national
command economy.

9800.indb 113 6/2/16 3:05 PM

114  Chapter 4

The associated costs and scale of such a supercharged system were

accordingly colossal. Glushkov captured the sentiment of network effects,
which is still alive in surveillance capitalism’s promotion of big data today,
in this phrase: “world practice shows that the larger the object for which
an information-management system is created, the greater its economic
effect.”8 More than komchamstvo, or Lenin’s term for “Communist boast-
ing,” the basic OGAS blueprint affirms its staggering magnitude. In its ini-
tial proposals, the OGAS Project estimated that it would take over thirty
years to be fully online, that it would need a labor transfer of some 300,000
personnel, that costs would be upward of 20 billion rubles for the first fif-
teen years, and that tens of thousands of computing center and interactive
access points would be distributed across the Soviet population.
All this would prove net efficient, promised Glushkov. The 300,000
knowledge workers would constitute an enormous labor transfer, as well as
a net reduction in the ever-rising number of people who were employed in
economic planning. The 20 billion rubles would be distributed over three
five-year plans, with the first requiring a seemingly modest 5 billion rubles.
Acutely aware of the advantages of the well-regulated financial management
that was enjoyed by the successful military nuclear and space programs,
Glushkov insisted to Prime Minister Kosygin that, if the OGAS were to be
developed, this civilian program would require a similarly well-managed
funding stream, even though it would prove more complicated and expen-
sive than both military programs combined. For his distinctly decentralized
civilian economic communication infrastructure project, Glushkov sought
fully centralized military-style financial funding. Only with well-managed
funding could this civilian project pay for itself, which it promised to do
handsomely, returning fivefold on the first fifteen-year investment, or “no
less than 100 billion rubles” (roughly $850 billion in 2016 U.S. dollars), and
even this estimated windfall in savings “was a conservative figure.”
Cost, in other words, is the simplest reason that the OGAS Project never
developed as proposed. A networked command economy, as economist crit-
ics noted, would simply prove uneconomical. No such sum of funding was
granted, and the projected costs soared slowly upward until, according to
varying estimates, the OGAS, if built in the late Soviet Union, would cost
the staggering sum of 160 billion rubles (or $1.4 trillion in 2016 dollars, or
roughly the U.S. deficit in 2009).9 Still, costs are never black or white. The
OGAS Project imagined a series of adjunct projects with less painful price tags.
As early as 1963, the EGSVT technical network proposed a far more afford-
able fraction of this vision—one center in Moscow, twenty to thirty regional
computing centers, and unspecified local computing “access points.”

9800.indb 114 6/2/16 3:05 PM

Staging the OGAS, 1962 to 1969  115

The Visionary behind the Vast Network: Viktor Glushkov

Viktor Glushkov (1923–1982), who was called the “king of Soviet cybernet-
ics” in his New York Times obituary, was neither the first nor the last to pro-
pose a nationwide network. But he figures as the organizing protagonist of
the remaining history as the leading champion of the OGAS Project, a well-
positioned academician, vice president of the Academy of Sciences, and a
leading cyberneticist. Known as both a global thinker and a local doer from a
young age, Viktor Mikhailovich Glushkov was born in the temperate south-
ern Russian city of Rostov-on-Don on August 23, 1923, into a family of a
mining engineer (figure 4.3). Like many prominent Soviet figures, he excelled
in mathematics at a young age and in middle school dreamed of becom-
ing a theoretical physicist. In high school, he quickly grasped topics such
as quantum mechanics and absorbed classics in the original German from
Johann von Goethe to Georg Wilhelm Friedrich Hegel’s The Philosophy of His-
tory. In 1941, the Nazis executed his mother for her part in the underground
resistance. After failing to enlist in the artillery school for health reasons, he
turned to mathematics in college, dove into topological algebra, and gradu-
ated in 1948. Four years later, including two years to complete his doctorate
while holding a research position at a new nuclear center in Yekaterinburg
(then Sverdlovsk) in central Russia, he proposed solutions to David Hilbert’s
generalized fifth problem in 1952. In 1900, in Paris, the German mathema-
tician David Hilbert proposed twenty-three foundational problems that
have attracted much attention in modern mathematics since. Two of those
problems are considered unresolvable, and the fifth problem, parts of which
Glushkov tackled, involves smooth manifolds in Lie group theory. The initial
breakthrough came to him while he was climbing an ice field on Mt. Kazbek
in the Caucasus with his wife, Valentina Mikhailovna. Six months later he
had formalized the shortest solution to that problem to that day.
This feat guaranteed that in the mid-1950s, the rising algebraist could
have secured almost any position in the Soviet Union. Thanks to an intro-
duction from academician Boris Vladimirovich Gnedenko, Glushkov
became acquainted with Lebedev’s computing center in Kiev, which six
years later (in 1962) he transformed into the prominent Institute of Cyber-
netics. He directed the institute from 1962 until his death in 1983. When
asked why he chose to shift his attention to the intersection of computer
technology and mathematics and subsequently assume the directorship of
a Computing Center from Sergei Lebedev in Kiev, Ukraine, in 1956—and
not a more politically prestigious position in Moscow—he is reported to
have replied that his wife, Valentina, whom he had met in their third year

9800.indb 115 6/2/16 3:05 PM

116  Chapter 4

Figure 4.3
Viktor Glushkov, about 1963. From the personal
archives of Viktor Mikhailovich Glushkov.

of college in relatively balmy Rostov, preferred the warmer weather in Kiev

and he agreed.10 It is also possible that this committed theorist of decentral-
ized power saw a position removed from Moscow as a strategic opportu-
nity to practice and leverage decentralized power. So having achieved an
ambitious goal in mathematics at a young age in 1956, Glushkov turned
his sights to theorizing the emergent field of cybernetics, especially the
relationships between information technology and economic cybernet-
ics. His oeuvre swept across theoretical fields (including abstract algebra,
mathematical logic, automata theory, and algorithms) and applied fields
(including the development of hardware, software, robotics, informatics,
and computers and the administration of Soviet economic cybernetics).
Glushkov is remembered by colleagues for having been always “on”—a
persistent kind of applied grand theorist—except for the occasional hike or
fishing trip down the Dnieper River, which he relished. His children recall
him following a strict daily regime: when he was not riding the day-long
Kiev-Moscow train (which he jokingly called his home), he rose at 8:30

9800.indb 116 6/2/16 3:05 PM

Staging the OGAS, 1962 to 1969  117

am, exercised, breakfasted, went to work, returned home in the evening,

and continued working until about 2:00 or 3:00 am. In 1963, as the first
director of the brand new Institute of Cybernetics, his work habits reached
a feverish pitch and then broke. Valentina recalled that he worked eigh-
teen to twenty hours a day until at age forty, he suddenly collapsed from
a brain seizure. (A tumor of the medulla likely ended his life twenty years
later.) Undeterred and still bound to his hospital bed, he finished the intro-
duction to his Lenin Prize–winning book, The Design of Digital Automatic
Machines. His intense persistence of mind rendered possible his mathemati-
cal achievements, and his vision was shortsighted since youth due to his
voracious reading habits. It is not known whether this contributed to his
protracted struggle with a fatal brain tumor.
Fluent enough in German and English to lecture and publish abroad in
those languages (having once recited excerpts from Goethe from memory
for two hours to win a bet), Glushkov figures as a consummate informa-
tion universalist, even among cyberneticists, for whom practically every
challenge reduced to, as his colleague and fellow computer pioneer Boris
Malinovsky put it, “the global problem of the computerization of informa-
tion sharing.”11 Committed to building computer networks that share infor-
mation, his subsequent research goals pushed him to generalize his applied
innovations further and further. Some characteristic examples include
career examinations in not just specialized computing but multipurpose
control computing; not just von Neumann computing processor architec-
ture but “massively parallel macro-piping” and a recursive base for frac-
tal processing in computer architecture; not just computer programming
but natural-language computer programming; not just robots but entirely
digital automata; not just bureaucracy but paperless offices and informat-
ics; and in the end, not just a better economic life for private humans and
our collective humankind but an even bolder and more remote future. He
identified in the inevitable evolution of artificial intelligence the possibility
of “informational immortality,” where the subjective consciousness, mem-
ories, and personalities of individuals and societies might be transferred
into a global network that was capable of outlasting the ages, resurrect-
ing and recasting civilization as we know it.12 Because they reach so far,
the endpoints of these various research initiatives begin to express in relief
the grander vision that organized his personal commitment to the OGAS
Project as the next step in networking onto the higher plane of the grand
collective of socialist labor.13 For Glushkov, the OGAS Project represented a
vehicle for achieving the whole of his many scalable visions.

9800.indb 117 6/2/16 3:05 PM

118  Chapter 4

Glushkov modeled his thinking about computer networks and process-

ing after—and often against—the prevailing trends in the study of neu-
ral networks. In a notable deviation from von Neumann digital computer
architecture that pushes all data bits through a bottleneck one bit at a time,
Glushkov theorized about what he called the “macropiping” or “macrocon-
veyor” processor architecture for transmitting information along multiple
processors simultaneously between groups of computers. Macropiping was
modeled after his cybernetic vision of the computer, which, according to a
1959 speech, would best resemble the human brain in its capacity to pro-
cess billions of bits of data in parallel simultaneity. This idea germinated
into his notion of a simultaneous national network that would function as
a self-regulating nervous system for the whole of the Soviet people. Glush-
kov shared conversations and computing technology with people such as
chessmaster Mikhail Botvinnik, among many other ambitious dreamers, to
create a machine in the image of man, not the other way around. In the
late 1950s, Glushkov sought to develop in theory a computer programming
that imitated the sophistication of human thought, cognitive function, and
natural language. For example, he and his colleagues examined processes
for distinguishing between grammatically and semantically correct sen-
tences, such as “The chair stood on the ceiling,” as a step toward achieving
natural language programming and a more human “higher intellect” in the
The OGAS Project took shape in a complex network of research teams
(at the center of which sat Glushkov). No science is a solitary endeavor,
however, and a full accounting of the details of the people who constituted
Glushkov’s teams, their accomplishments, and their frustrations is beyond
the scope of this book. Two of his favorite students and eventually a wife-
husband team, Yulia Kapitonova and Aleksandr Letichevsky, identify what
they call the intellectual “school” of Viktor Glushkov, which itself con-
tained many teams that contributed to the OGAS Project and many other
projects. The first EGSVT proposal began to take shape in the conversa-
tions of Glushkov, Vladimir S. Mikhalevich (who directed the Institute after
Glushkov), Anatoly Kitov, A. Nikitin, and others, and the first government
document published on the EGSVT, on May 21, 1963, also highlights as
coauthors Anatoly Kitov, V. Purgachev, Yu. Chernyak, M. Popov, among
others. Key members and colleagues at the Institute of Cybernetics in Kiev
included Vladimir S. Mikhalevich, V. I. Skurikhin, A. A. Morozov, Yulia V.
Kapitonova, Aleksandr A. Letichevsky, A. A. Stognii, T. P. Mar’yanovich,
and others. The Moscow-based supporting scientists included Anatoly
Kitov, Yu. A. Antipov, I. A. Danil’chenko, Yu. A. Mikheev, R. A. Mikheeva,

9800.indb 118 6/2/16 3:05 PM

Staging the OGAS, 1962 to 1969  119

among others.15 Optimization modeling, which would have contributed

to the management software running the OGAS economic reform) were
developed from 1962 to 1969 by Vladimir Mikhalevich, O. O. Bakaev, Yu.
M. Ermol’ev, I. V. Sergienko, V. L. Volkovich, B. M. Pshenychniyi, V. V.
Shkurba, N. Z. Shor, and others. Glushkov toiled alongside A. A. Stognii
and A. G. Kukharchuk as principal designer in developing the Dnepr-2, a
transistor computer. He also headed a team that included Y. Blagoveshen-
sky, Aleksandr A. Letichevsky, V. Losev, I. Mochanov, S. Pogrebinsky, and
A. A. Stognii in developing the MIR-1 engineering calculation machine, an
exhibition version of which IBM purchased in London.16
Other supporting teams in the Glushkov school indirectly reflect on the
OGAS Project. Kapitonova and Letichevsky, for example, helped Glushkov
theorize an “analytic” mathematical human language programming lan-
guage and an algorithmic design in computer design automation.17 This
team helped nudge the field of artificial intelligence away from the notion
that the brain was machine-inspired (away from McCulloch and Pitts’s
claim that the brain follows logical circuitry). Instead, they worked on
building a brain-inspired machine that was “capable of carrying out com-
plex creative activities,” continuously seeking to reveal the “higher intel-
lect” of machines modeled after mechanisms of the mind.18 If there was a
danger in the brain and machine metaphor, it ran only one way for Glush-
kov: “the danger is not that machines will begin to think like people,” he
intoned, “but that people will begin to think like machines.”19

Are National Networks More Like Brains or Nervous Systems?

In 1962, Glushkov imagined the OGAS as a “brainlike” (mozgopodnobyi)

network for managing the national economy and extending the life experi-
ence of the nation and its inhabitants. Consider the implications for the
cybernetic analog between neural networks and national computer net-
works. As already noted, cybernetics brings to bear powerful conceptual
frameworks for imagining structural analogies between ontologically differ-
ent information systems—organisms, machines, societies, and others. The
cybernetic instinct rushes many visionaries to profound structural insights
but also to overly determined design decisions. The circuitry of a com-
puter chip and the neural networks of a mind do not resemble each other,
although cybernetics earns its keep by finding usable analogs between
them. This cybernetic system analog instinct—to design in beautiful sym-
metry where not necessary—helps to explain the consistent hierarchically

9800.indb 119 6/2/16 3:05 PM

120  Chapter 4

decentralized design of all Soviet national network projects. They were

designed to resemble the national economy as it appeared in principle, not
as it worked in practice. To quote the secretary of the history of the Central
Economic Mathematical Institute, a collegial institute of Glushkov’s Insti-
tute of Cybernetics, the decision was made to “build the country’s unified
net hierarchically—just as the economy was planned in those days.”20
In other words, Kitov, Glushkov, Fedorenko, and others followed the
cybernetic integration of machines and biology to its design conclusions.
Like Kharkevich’s design, Glushkov’s OGAS and other Soviet economic
cyberneticists insisted that the Soviet economy, as a national body, needed
a central information processor, administrator, and brain. They were not
alone in modeling national networks as a neural network in the early
1960s.21 The U.S. network engineer Paul Baran envisioned the ARPANET as
a distributed packet-switching network that was modeled in part after War-
ren McCulloch’s vision of the brain.
Note the difference here: Soviet economic cyberneticists under Glushkov
conceived of the national network as a match for a national economic body
with the network as the nervous system complete with a central processing
in Moscow, and the American model of distributed networking imagined
the whole of the nationwide computer network after the dynamic struc-
ture of the brain itself, not the body. In the Soviet Union, the command
economy resembled the body, with the economic planning apparatus as its
nervous system and Moscow planners as the brain, and in the West, after
the ARPANET was commercialized, there was no body outside of the brain
itself: the whole national network of users made up the nationally distrib-
uted brain itself.
To reduce it to a simplistic cold war binary: cybernetic network entre-
preneurs throughout the world had competing analogs for thinking about
national networks. In America, the ARPANET was designed to resemble a
brain of the nation because its visionaries first imagined the nation as a sin-
gle distributed brain of users. In the Soviet Union, the OGAS was designed
to resemble a nervous system for the nation because its visionaries first
imagined the nation as a single incorporated body of workers. This Soviet
analog between network and nervous system, far from determining the
outcome of the network, also occurred in Project Cybersyn in the early
1970s in Chile. Its principal architect, the British cyberneticist Stafford Beer,
sketched the socialist Salvador Allende’s nation as a viable system that was
based on the “human nervous system” analogized with a comprehensive
firm or corporate organization—complete with executives in adaptive feed-
back loops with the national body of workers.22

9800.indb 120 6/2/16 3:05 PM

Staging the OGAS, 1962 to 1969  121

In addition to taking the cybernetic brain-computer analogy to its logical

extreme, Glushkov also sought tight structural analogies in the communica-
tion systems that connected technical and human machines. For example,
he designed the programming language Analytic to resemble human speech:
“We continued to develop it in accordance with the principles of progres-
sively complex machine languages, to get closer to human language.… My
goal was to be able to speak directly with the computer and issue commands
in our language.”23 Like the relationship between neural, processor, and
national economic networks, Glushkov’s thoughts about scripting together
natural language and computer programming rests on the assumption that
there is nothing particularly natural about natural language and that com-
puting coding (like his other conceptual innovations in macropiping pro-
cessing, automata, and the paperless office) represented an extension of the
calculable artifice already hard at work in human behavior.24
Each of his innovations sought to reframe and solve knotty local prob-
lems in terms that scaled to a larger global system that contained those
problems and all those like them. In fact central to understanding Glush-
kov’s life and work and his scalable vision for the OGAS is his unflagging
intellectual commitment to what he called “practical universals”—the
merging of mathematics and economics, the theoretical and the applied,
the universal and the particular. He and his colleagues repeatedly insisted
that three principles guided his life work—“the unity of theory and practice,
the unity of distant and near goals, and the decentralization of responsibil-
ity.”25 He taught others that before putting a principle into action, they had
to formulate it into a general model or rule in abstract mathematical terms
and then test that rule practically, applying it to countless concrete exam-
ples—an imperative to act locally while thinking globally. When higher
authorities handed six of his researchers seven discrete system problems,
Glushkov insisted that the first step was to develop a universal language
for modeling all discrete systems, a language by which they could then
solve all seven problems simultaneously, as well as any more they could be
given.26 The OGAS in design and implementation followed suit: people at
each step of the network—including factory-based control system, regional
computer center, and national economic planning center—sought to solve
short-term factory problems by developing a universal system for advanc-
ing Soviet socialism toward communism.
The OGAS, for Glushkov, was to be a national communication network,
countless local paperless offices, and a dynamic management system that
connected them—a global-local network. A proper economic reform, in his
mind, must benefit the factory worker, the general secretary, and the whole

9800.indb 121 6/2/16 3:05 PM

122  Chapter 4

populace. The OGAS sought to pole-vault socialism toward communism at

the Hegelian level of historical progress and to usher in a better work life
for the knowledge worker: in the command economy, everyone needed to
work knowledgably with economic plans. The OGAS would grant both at
once, automatically storing relevant digital files on every local actor while
granting remote access anywhere else in the country. The origins of the
ideas behind the OGAS computing network also point to a preexisting aca-
demic network, including the circulation of a 1955 Academy of Sciences
proposal by Nemchinov to erect large but unconnected state computer cen-
ters in Moscow, Kiev, Novosibirsk, Riga, Kharkov, and other major cities.
However, this proposal did not connect the computer centers but instead
specified that they should be built to facilitate the local exchange and stan-
dardization of scientific and economic information. Kitov’s Red Book letter
in the fall of 1959, which included the initial proposal to network such
computers together into one, was the next step. In fact, after Kitov was dis-
missed from the military, Glushkov hired him to serve as a scientific adviser
and personal confidant to his projects. Their respective trust network grew
so close that, two decades later, one of Kitov’s sons and one of Glushkov’s
daughters wed, signifying, just as the close connections between Baran and
McCulloch, that personal communication networks both precede and out-
last national computing networks.27
Beginning in the early 1960s, Glushkov’s detractors recognized the
sweeping commitment to practical universals in this vision and colored
it in different shades. As he exercised his penetrating ability to formulate
and scale up or down any problem by the force of mathematical reason,
Glushkov’s vision of the socialist cybernetic future moved, in the estima-
tion of researchers at the Central Economic-Mathematical Institute (CEMI)
and liberal economists, in “romantic” and “quixotic” leaps. Even his col-
leagues admitted in interviews that at the grandest vision, the OGAS ambi-
tion had an almost “religious” or cosmological reach to it.28 The modern
reader should suspend incredulity at the scope of his theoretical scale until
after observing the similar scale of technological ambitions at work else-
where. The totalizing corporate missions of modern-day major data com-
panies and the scope with which data are harvested by corporations and
states share intellectual affinities with the all-inclusiveness of his or any
global-local network vision. Glushkov was not alone in 1963 in proposing
that the state should gather dossiers on every worker and economic actor
in his nation.
By contrast, Glushkov’s proponents, caught up in both the breadth and
precision of his plans, too often overlooked the frequent criticism that no

9800.indb 122 6/2/16 3:05 PM

Staging the OGAS, 1962 to 1969  123

institutional environment could possibly be ready to do all that the OGAS

sought to do. Glushkov also recognized that no practical effort, no matter
how impressive, could ever satisfy both the local and global demands of
making paperless the command economy, and many of his career efforts
outside the OGAS Project focused, to his credit, on local projects, including
the paperless office.29 For the OGAS, however, because it was a matter of
economic bureaucracy reform, he insisted on a comprehensive meaning of
economic information: “since the object of control is not only equipment
but also personnel, one must include [in the OGAS] all the information
about new technical, technological, economic, and organizational ideas
and projects that workers at a given enterprise have.”30 Far more than a
shared file containing economic information, the OGAS presented itself
as a real-time clearinghouse for information concerning individuals, proj-
ects, factories, enterprises, and industries. The network would continue to
expand in scope, according to Glushkov, until it encompassed the whole of
the Soviet economy as well as all workers, their activities, and their office
space. At worst, the vision appears a totalizing and decentralized (not totali-
tarian) information capture of the workers and their work environment.
At best, it appears to be an organization information upgrade that is fit for
every large-scale corporation. Depending on how one weighs the values
of individual privacy and organizational purpose, these two champion a
particular universal ethical tension that occupies the modern media age.
Glushkov also foresaw (or rather projected) a hint of the financial future,
although perhaps not the future he had hoped for. Because the socialist
economy would be incrementally organized into a cybernetically balanced
network of labor, production, and consumption inputs and outputs, Glush-
kov reasoned, there would remain no reason not to virtualize currency itself
and make the exchange of funds take place by “electronic receipt.” With
the OGAS operational, there would no longer be need for hard currency.
All economic exchanges would take place online. Following this line of
thought, Glushkov included in his initial OGAS draft proposal a notewor-
thy provision to eliminate all paper currency, providing in its place wire-
less money transfers, or a “moneyless system of receipts” over the OGAS
network.31 Although modern readers may be tempted to see in his proposal
a prototype of the modern-day ATM, e-commerce digital money transfers,
PayPal, or BitCoin, Glushkov framed paperless money transfers in the poli-
tics of his time and place, calling it the fulfillment of a Marxian prophecy of
a future Communist society without hard currency. Read backward in pre-
sentist terms, as historians are loath to do, the proposal, if realized, would
have transformed the Soviet Union into, in Vladislav Zubok’s phrase, “a

9800.indb 123 6/2/16 3:05 PM

124  Chapter 4

computerized socialist utopia, the motherland of the Internet and also pos-
sibly the ATM.”32
I maintain that the historical lesson is that whatever our present-day
language and whatever the future imaginations of hard currency, the past
brims with a variety of visionaries who thought about the future of money
as virtual, when as history instructs, the dominant form of currency has
already always been, since ancient Mesopotomia, the arithmetic matter of
credit and debit—itself a form of expectant funds, or money transfers made
virtual across time, not space.33 After reviewing the proposal, Keldysh, then
president of the Soviet Academy of Science and a major supporter of Glush-
kov, asked to meet with Glushkov privately and urged Glushkov to strike
from his original OGAS proposal the recommendation of a networked soci-
ety without hard currency out of fear that it would raise “unneeded emo-
tions.” He warned Glushkov that the reviewing Soviet administrators were
so deeply attached to the advantages of hard currency that no reasoning or
ideological commitment could persuade them to abandon it.34 Glushkov
conceded the point, and the Central Committee initially approved his proj-
ect, pending further review.

Glushkov as a Pragmatic Administrator

In many ways, Glushkov, whatever his sweeping visions in cybernetic the-

ory, proved to be a pragmatic administrator in practice. Because he under-
stood practical administration, he also knew that the inevitable limitations
of his theoretical ambitions were, in fact, an important part and conse-
quence of his approach to problem solving with practical universals. If the
Soviet administrative system worked informally behind the scenes, then
so must he, especially if he wanted to help to rationalize or formalize that
same system. Unlike some of the stillborn or short-lived cybernetic propos-
als noted earlier, the longevity of the OGAS as a potentially viable network
project owes a debt to the tenacious and pragmatic administrative acumen
that Glushkov and his colleagues displayed in navigating, managing, and
alliance forging in the administrative base of the Soviet state between 1962
and 1983. As illustrated by the Komsomol letter incident and repeated by
many of his colleagues, Glushkov was sensitive to the political nature of
the OGAS proposals and, with his upper-echelon supporters, strategically
planned every step of coalition building around every part of his proposal:
who would support what and why.35 No naïve technocrat, he sought to
shape and situate his proposal according to the governing logics of blat and
personal politics.

9800.indb 124 6/2/16 3:05 PM

Staging the OGAS, 1962 to 1969  125

His wireless currency proposal is a case in point. On Keldysh’s advice, he

promptly dropped the idea and turned his attention back to the practical
universals that he would need to understand before he could integrate both
macroeconomic designs and microlevel problems of the Soviet economy.
Glushkov sought out and marinated himself in the practices of the actual
command economy so that he would understand locally what he sought
to reform universally. In the early 1960s, Glushkov received permission
from Keldysh to observe how each of the constituent parts in the Soviet
economy—factories, firms, collectivized farms of all types, and administra-
tive organs like local, regional, and national planning committees—actu-
ally worked. His purpose was ethnographic—“to ask questions, or simply sit
in the corner and watch how they work: what he decides, how he decides
it, according to what principles, etc.” Glushkov recalls, “And naturally I
received permission to acquaint myself with any industrial object—corpo-
rations, organizations—that I wanted.”36 By 1963, Glushkov reported hav-
ing visited and observed over a hundred such industrial sites and nearly a
thousand over the next decade, including mines, kolkhozes (collective farms),
sovkhozes (Soviet state farms), railways, an airport, higher control organs, and
administrative organs at Gosplan (the Soviet ministry charged with planning
the Soviet economy) and the Ministry of Finance. Glushkov claimed that “I
may know the structure of the national economy better than anyone else:
from the bottom up, I know the peculiarities of the existing controls system,
the difficulties which occur, and the most important issues.”37
At roughly the same time that the OGAS proposal was being reviewed
by the Central Statistical Administration, Glushkov gained insights into
the navigation of the informal complaint culture and the administrative
mechanisms that were available for resolving them. Between 1966 and
1976, he served as a Kiev-based adviser for the Division of Clemency (otdel’
pomilovaniya) for the prominent city of Kharkiv in eastern Ukraine. His
behavior in this public function also provides a glimpse into his adminis-
trative behavior concerning top-secret projects such as the OGAS. In these
archival materials, a pattern emerges. For each complaint case that he con-
sidered, he sent a formal letter and an informal letter. The first letter he
sent to the complainant to offer his moral support but declare his likely
inability to ease their situation, and the second letter he sent to the relevant
supervisory institution pleading informally the strongest appropriate case
on behalf of the complainant. Thus he resolved dozens of real-life con-
flicts within the actual social economy of formal appeals and complaints,
including helping a grandmother campaign against alcoholism, speeding
a mother’s request for an apartment, acquitting a decorated war veteran

9800.indb 125 6/2/16 3:05 PM

126  Chapter 4

convicted of unspecified crimes, and restoring to his studies a graduate stu-

dent found guilty of “hooliganry” for being found in a “nonsober” condi-
tion.38 At the same time that he was navigating this public trading zone
between the superabundant conflicts of bureaucratic and real-life inter-
ests, he was developing the OGAS as a top-secret human-computer system
proposal that would do the same—resolve informal conflicts at a national
economic level. Next I look at how the informal behind-the-scenes work
culture of these cyberneticists contextualizes this larger point.

“Cybertonia”: From National Cyberculture to Local Counterculture

Glushkov’s proposal to rationalize and automate the national economy

in 1962 took shape just as his own institutional environment was being
upgraded from a small computing center to a more ambitious formal set-
ting of an academic institute, without losing its informal and, in after-work
hours, almost countercultural work environment. In the early 1960s, his
vision for reforming the command economy took on national ambitions at
the same time that his own local institution entered national prominence.
A glance at the institutional transition from Sergei Lebedev’s laboratory in
the valley of Feofania to Glushkov’s Institute of Cybernetics will provide
insights into how the local institutional culture of this particular transition
animated both formal and informal attempts to imagine an alternate Soviet
information society.
The formal history of the transition from computing center to academic
institute is illustrious if not unusual. In the late 1940s and 1950s, Sergei
Alexeyevich Lebedev gathered a small and extraordinarily talented group of
electrical engineers into a computing laboratory in the valley of Feofania in
the southern outskirts of Kiev, Ukraine. That small group brought into exis-
tence the MESM (malaya electronicheskaya schetnaya mashina, or the small
electronic calculating machine, and predecessor to the mainframe work-
horse BESM series), the first stored-memory electronic computer in Europe,
arriving four years after von Neumann’s UNIAC39 (figure 4.4). In 1952, the
first “large electronic computer,” the BESM, or bol’shaya electronicheskaya
schetnaya mashina, followed, and then a series of Soviet native mainframe
computers—the M-20, the BESM-3M, BESM-4, M-220, M-222, and finally
the BESM-6. Designed in 1966 and produced first in 1968, the impressive
BESM-6 went into serial production and served in special-purpose compu-
tation centers and military computer networks for the next two decades. In
1962, under Glushkov’s direction, Lebedev’s laboratory was relocated a mile
away to a separate campus facility of the future Institute of Cybernetics that

9800.indb 126 6/2/16 3:05 PM

Staging the OGAS, 1962 to 1969  127

Figure 4.4
The MESM (small electronic calculating machine) and its team in the monastery near
the cathedral in Theophania, 1952.

was known for a series of subsequent impressive achievements. Researchers

at that facility developed the “Dnieper” computer series, which powered
the base stations for Soviet cosmonaut flight south of Moscow while press-
ing the frontiers of Soviet information science and technology. The insti-
tute also is known for developing the mainframe and early microcomputers
Mir and Promin and a range of research on economic cybernetics, medical
cybernetics, artificial intelligence, optimization, and defense research. The
projects included the first network project to digitize the entire command
economy and their central project—the OGAS and its technical base EGSVT
beginning in 1963. In all, the official histories convey the gravitas that one
would expect from one of the elite teams of Soviet scientists.
A closer look at the local practices of these institutions, however, sheds a
very different light on this moment of Soviet optimism. The years 1962 and
1963 marked the height of enthusiasm for a young, entrepreneurial, and
surprisingly humorous and mischievous group of cyberneticists. Lebedev’s
laboratory was situated in a forest that was enchanted with Slavic legends.
Overrun by songbirds, rabbits, mushrooms, and berries in the summer and
haunted in the winter by rumors of wolves and Baba Yaga (the famous

9800.indb 127 6/2/16 3:05 PM

128  Chapter 4

witch of eastern European folklore), this forest served as a curiously natu-

ralistic cradle for Lebedev’s MESM, which was then the emblem of the new
Soviet religion of rational scientific progress. In the center of an opening in
the woods stands St. Panteleimon’s Cathedral (Panteleimonivs’kii sobor), a
high point of Russian revival ecclesiastical architecture since its construc-
tion in 1905 to 1912 (figures 4.5 and 4.6).
Nearby stands a two-story brick building that tells a story of a compli-
cated intersection of faith, madness, murder, and science. Initially built as a
dormitory for Eastern Orthodox priests, the building was looted during the
1917 Russian revolution and converted into a psychiatric hospital. In 1941,
the Nazis murdered its patients and established it as a military hospital. In
1948, the badly damaged building was transferred to Lebedev’s work on the
newest icon of Soviet atheism—that triumph of human rationality and cre-
ativity that was the automated computer. Six thousand vacuum tubes and
two years of astonishing effort later, Lebedev’s team turned on the monster
calculating machine in 1950. A sense of collaborative, dedicated work ethic
lingered in the decades thereafter, and a sense of local autonomy that was
away from the watchful eyes of Moscow pervaded the area of Feofania.
Researchers who received housing nearby rarely chose to leave, even when
offered more prestigious positions. Informal play and even troublemak-
ing abounded. To the priests’ chagrin today, engineers sometimes tested

Figure 4.5
St. Panteleimon’s cathedral and monastery (left), which housed the MESM.

9800.indb 128 6/2/16 3:05 PM

Staging the OGAS, 1962 to 1969  129

Figure 4.6
Park, pond, and forest in Feofania, the general setting for Sergei Lebedev’s computing
laboratory, late 1940s to 1950s.

controlled mechanical explosions in the magisterial monastery. Water

fetched from a nearby well was used to extinguish the fires because the
building where the first computer in Europe was built had no plumbing.
After work, the mood lightened. Bus drivers were sent on wild goose chases
through the forest, and juggling and ping-pong balls ricocheted down the
hallways of offices and laboratories. On work breaks, volleyball and soccer
games broke out, and after work, the researchers ran to swim in the nearby
lake and to wander through the tall pines and oak trees of the surrounding
forest. Lebedev and Glushkov are rumored to have drafted the organization
of the Institute of Cybernetics, built three kilometers to the west, while
strolling together through that forest.
When the Academy of Sciences appointed Glushkov to be the first
director of the new Institute of Cybernetics, some of that informal spirit
transferred to the new institution, in part thanks to a prolonged transition
period during the 1960s in which the campus where the institute is cur-
rently housed was built. In the after-work hours and at holiday parties, the

9800.indb 129 6/2/16 3:05 PM

130  Chapter 4

growing group of young institute researchers even imagined a humorous

autonomous country of their own, “Cybertonia,” a virtual country. The
researchers, whose average age was roughly twenty-five, first christened
this “fairytale [skazochnaya] land” during a New Year’s Eve party in 1960.
The joke snowballed. The fairytale land offered scientific seminars, lectures,
films, and auctions mainly in the capital Kiev and an evening ball in the
Ukrainian nationalist border city of L’vov, spinning off more and more
activities (artwork, ballads, a short film, passports, currency), press releases,
seminars, holiday and after-hour gatherings, community functions, and
more parties.40 The researchers at the Institute of Cybernetics were still sev-
eral years away from occupying the Institute’s future campus, which even-
tually included more than a dozen buildings along Glushkov Prospect in
southwest Kiev (figure 4.7). From 1962 to 1970, the institute occupied a
building at 4 Lysogorskaya Street several kilometers north, at an intersec-
tion with Nauka (Science) Street, an area famous for being featured in the
science fiction of the Strugatskii brothers, who worked in the Institute of
Physics a few blocks away41 (figure 4.8).
In its informal practices, the Cybertonia society abounded in pranks,
puns, and puzzling wit, recreating a country in the image of the autono-
mous Soviet automata. The collective issued fake stamped passports and

Figure 4.7
Sketch of the Institute of Cybernetics campus, Prospect Academic Glushkov, 40,
Kiev, 1970.

9800.indb 130 6/2/16 3:05 PM

Staging the OGAS, 1962 to 1969  131

marriage certificates to the mostly male research staff and female adminis-
trative staff, authorized by the “Robot Council of Cybertonia.”42 (figures 4.9
and 4.10). Each passport packed mathematical equations into the blanks for
personal identification, accompanied by a national constitution and a map
of the future capital of “Cyber City” (Kybergrad). The workplace culture
at this prominent research institute embraced the joke as an ambiguous
means for letting a little steam off after work and, in their more ambitious
flights of imagination, envisioning a nation that was independent from the
Soviet Union. The blurring of reality and virtuality, work and play, science
and art was the point of “Cybertonia,” a name that lives on in the title of
an academic journal recently begun by Glushkov’s youngest daughter, Vera
Viktorevna Glushkova.43 The Cybertonia constitution guaranteed the rights
to frivolity and humor complete with the faux-newspeak warning: “anyone
who disobeys the Robot will be stripped of their rights and cast out of the
country for 24 seconds” (figure 4.11). The map featured landmarks such as
“a Main Post Office and the Feedback Division (or Returned Communica-
tion),” or Glavpochtamt y otdel obratnyi svazi, a possible reference to Cyber-
tonia as a self-contained system apart from the Soviet regime, as well as the
“Temple of the 12 Abends” (abnormal program ends, or software termina-
tions), or Khram 12 avostov, a near Russian homophone with “the Temple
of the Twelve Apostles.” Currency was issued on the punch cards that were
used in analog computer memory storage.
Perhaps most boldly, the Cybertonia society hosted a saxophone-playing
robot mascot as a unveiled reference to jazz, an export of American global

Figure 4.8
Sketch of the Institute of Cybernetics building, Lysogarskaya 4, Kiev, 1966.

9800.indb 131 6/2/16 3:05 PM

132  Chapter 4

Figure 4.9
Cybertonia passport, 1965. (a) front, (b) back.

9800.indb 132 6/2/16 3:05 PM

Staging the OGAS, 1962 to 1969  133

Figure 4.10
Cybertonia wedding certificate, 1965.

Figure 4.11
Constitution of the country of Cybertonia, about 1966.

9800.indb 133 6/2/16 3:05 PM

134  Chapter 4

culture (figure 4.12),44 and it published at least one issue of a newspaper

and made a comedic short film titled “Feofan Stepanovich serditsya” (fig-
ure 4.13). By 1966, its motto had evolved to “energy, laughter, dreams,
and fantasy.” Stamped on the headline of the single issue of the group’s
newspaper the Evening Cyber stood the greetings “s novyim kodom” (or
“happy new code,” a near homophone with “happy new year” in Russian).
In 1968, a season ripe with revolt, a symposium of cybertonians published
an irreverent report on the “complex cybernetic aspects of humor” that was
issued from “Cyber City” in April 1969. The report contains nothing explic-
itly subversive but overflows with technocratic wit and sarcasm directed
against Soviet authority figures. These merry pranksters compared the task
of securing living quarters (a notorious challenge of everyday Soviet life) to
hyperdimensional geometry and published “formal” reports on “theory of
Graphs/Counts” (teoriya grafov, the royal title of count is a homophone with
the word graph in Russian), a Jonathan Swift–like account of laughter at
work as an underutilized national economic resource, odes to the virtues of
Georgian soccer, cheese, beer, and a few chauvinistic laughs about the pros-
pects of the feminization of science. Another report in 1965 bore the bold

Figure 4.12
Cybertonia logo: a robot playing jazz on a saxophone, about 1966.

9800.indb 134 6/2/16 3:05 PM

Staging the OGAS, 1962 to 1969  135

Figure 4.13
Parody newsletter: Vechernii Kiber (Evening Cyber), 1966.

title “Executives Incognito: On Wanting to Remain Unknown, at Least to

the Authorities.”45 Puns punctuated the technocratic discourse while qui-
etly resisting power. These scientists sought in Cybertonia their own Cybe-
ria away from Siberia, an escape from the great error of Khrushchev’s age if
not the great terror of Stalin’s. Alas, Cybertonia never did grow to become,
as the editors of its 1968 symposium had gleefully enthused, an “interplan-
etary congress.” At some point between 1969 and 1970, as the Brezhnev
doctrine compelled the Warsaw Pact to invade Czechoslovakia, “the entire
idea of Cybertonia,” as a participant recalled, “was buried by the pressure of
the Party and government.”46
The purpose of this snapshot into the informal lives of Soviet cyberneti-
cists should be clear. In the forests of Feofania and in the virtual playground
of Cybertonia, network entrepreneurs sought intellectual, political, and
social autonomy, revelry, and even subtle informal protest from the oppres-
sive regime that they served. Just as other cultures have demonstrated the
rich connections between informal countercultures and cybercultures,

9800.indb 135 6/2/16 3:05 PM

136  Chapter 4

lively network forums reproduced the cultural, institutional, and gendered

mores of the Soviet 1960s, conceiving of a kind of privileged cybercom-
mune of their own making.47
In the early 1960s—when Glushkov’s ambitious plan to network, account
for, and automate the nationwide command economy faced both partial
formal approval and informal resistance from the top state authorities—his
own local institution was undergoing significant institutional growth even
as it was being told it must develop the EGSVTs before the OGAS network.
In this fleeting period of optimism, the establishment and growth of the
Institute of Cybernetics led to a form of institutional adolescence in which
it exercised institutional ambitions on the national stage while informally
and internally venting a kind of countercultural defiance against the state
regime that governed it.
In fact, at the same time, 1962 to 1968, that Cybertonia was being cel-
ebrated during after-work hours, the Institute of Cybernetics was transi-
tioning from a relatively small set of buildings near Theofania to a spacious
campus a few kilometers to the southwest. It had enough modern buildings
to house each major field of cybernetics with its own research department
(except for Glushkov’s “theoretical and economic cybernetics,” which
remained a department that preserves to this day the particular universal of
Glushkov’s merger of mathematics and economics).

CEMI and the OGAS Institutional Landscape in the 1960s

Glushkov’s research institute was not alone in experiencing institutional

growth in the early 1960s. Many prominent research institutes were estab-
lished across the Soviet Union in the 1960s (Ukraine today has roughly
130 research institutes, and Russia has many more). Under the leadership
of Aksel’ Berg and the new president of the Soviet Academy of Sciences,
Mstislav Keldysh, most of these pertained to cybernetic research. Those
focusing on economic cybernetics included Viktor Glushkov’s Institute of
Cybernetics in Kiev and Nikolai Fedorenko’s Central Economic-Mathemat-
ical Institute in Moscow.
These academic institutes were located in the capitals of the Soviet empire
and under the umbrella of the Soviet Academy of Sciences. They functioned
not as “islands of autonomy” (as may have been the case in the secret Sibe-
rian science city of Akademgorodok) but initially as contingent trading
zones and eventually holding stations for enthusiastic young researchers
who powered much of the early wave of Soviet cybernetic research growth
throughout the 1960s. Prominent research institutes of all kinds sought to

9800.indb 136 6/2/16 3:05 PM

Staging the OGAS, 1962 to 1969  137

establish a rolling range of connections, although they often were prohib-

ited from doing so in lasting ways. In many cases, the most crucial alliances
and associations for the survival and success of their core research projects
rested on currying productive relationships with the governing state min-
istries, not peer research institutes, whose areas of responsibility affected
their research missions. The CEMI in Moscow, for example, effectively
became an operations arm for Gosplan and other large ministries, and the
Institute of Cybernetics in Kiev maintained greater degrees of separation.
The history of how these institutional alliances unfolded is the short his-
tory of the OGAS Project and its undoing. Some attention will be paid in
the following sections to outlining the formation and deformation of the
alliances between economic cybernetic research institutes and Gosplan, the
Ministry of Finance, the Central Statistical Administration, and the Minis-
try of Defense.
In 1963, Glushkov’s Institute of Cybernetics and another new power-
ful economics institute—the Central Economic-Mathematical Institute
(CEMI)—formed an alliance to advance the OGAS project, although the
seeds had been planted several years earlier. When Vasily Sergeevich Nem-
chinov—a senior economist-mathematician who was a strong advocate of
economic cybernetic reform and who had done much to introduce Kan-
torovich’s linear modeling and input-output mathematical models into
Soviet economic planning—was proposing the CEMI in 1960, he initially
called it the Institute of Economic Cybernetics and devoted it to Glushkov’s
main task of networking the national economy.48 The founding of CEMI
receives a moment of attention, too, because both new institutes invested
hundreds of young researchers and dedicated funding streams into devel-
oping the OGAS project.
Before CEMI was an institute, it was a small laboratory in Moscow in
1958 called the Laboratory of Economical Mathematical Methods. Nem-
chinov appealed to the Ministry of Finances of the USSR by letter in Janu-
ary 1962, claiming that the transformation of the Soviet economy from
socialism to communism depended on “optimal plans for the nation’s
economy.”49 By “plans” he had in mind the “optimal planning” of Kanto-
rovich’s linear programming as understood as both local microeconomic
modeling (which could be done on a standalone mainframe computer) and
a macroeconomic national infrastructure for processing the planned econ-
omy’s commands by computer. Initially inspired by Kitov’s failed 1959 Red
Book letter, Nemchinov appealed to the Ministry of Finance that “the mod-
ern mathematical methods and the means of mechanization and automa-
tion” were necessary to manage the complexity of the economy, invoking

9800.indb 137 6/2/16 3:05 PM

138  Chapter 4

Keldysh’s call in 1962 for “the transformation of economics into an exact

science in the full sense of the word.”50 Keldysh, the president of the Acad-
emy of Sciences of the USSR as of 1961, officially approved and promoted
economic cybernetic research as a priority of the academy, underscoring
that “the development of a theory of optimal planning and management
to a unified mathematical model of national economy was one of the main
directions of developments in modern economic science.”51
Nemchinov also employed cold war rhetoric to provide a sense of
urgency to his promotion of economic cybernetic methods as a means for
governing a society, socialist and capitalist alike, noting the strong similari-
ties between neoclassical econometrics in market economies and economic
cybernetics in socialist command economies: “after World War II these
methods were reopened in the West and were applied extremely widely to
monopolistic government planning.”52 He then invoked a sort of cold war
cybernetic economics gap, worrying that the Soviets had lagged behind the
use of cybernetic methods “in the internal planning of the most developed
capitalistic countries.”53
Yuri N. Gavrilets joined Nemchinov’s laboratory in 1959 and continues
to work at CEMI to this day. A former rocket engineer, he was one among
many military engineers who, like Kitov, was forced to pursue nonmilitary
economic research after a youthful display of what was interpreted to be
anti-Stalin activities in the late 1950s. According to an interview with him,
the early efforts of CEMI to incorporate mathematical methods into opti-
mal planning of the command economy openly sought to merge the best
Marxist principles of social justice and planning with capitalist free-market
equilibria.54 Inflating the threat of capitalist cyberneticists, Nemchinov con-
tended in his original proposal that “not a single scientific point” (which
he later crossed out by pen and replaced with the word center) currently
stood ready to “guide and coordinate research in the field [of economic
cybernetics]” in all of the Soviet Union.55 The Central Economic-Mathe-
matical Institute, Nemchinov concluded, together with the OGAS and its
associated mission of planning the economy by a cybernetic management
network, would fill just such a gap.
Nemchinov drafted his CEMI proposal several days after First Secretary
Nikita Khrushchev spoke at the Twenty-second Congress of the Communist
Party on October 18, 1961. That “secret speech” is remembered today for
denouncing the cult of personality, although Khrushchev also countered in
it Stalin’s bias against mathematical economics policy: “life itself requires
a much higher class of scientific foundations and economic accounts
from the planning and national economic leadership.”56 Cybernetics

9800.indb 138 6/2/16 3:05 PM

Staging the OGAS, 1962 to 1969  139

features frequently in Nemchinov’s official explanation of CEMI’s research

tasks, such as “the wide application of cybernetics, electronic calculating
machines and the regulating devices in production processes of industry,
construction industry and transport, in scientific research, in the planning
and project construction of calculations, in the sphere of accounting and
In a letter dated November 17, 1961, a month after the Twenty-sec-
ond Congress of the Communist Party, Nemchinov named four institute
research directives that were dedicated to the network vision laid out in
Kitov’s 1959 letter and Glushkov’s subsequent formulation of the OGAS:
1. The development of a unified system of planned economic information to im-
prove planned information and documentation companies, including work on the
application of modern calculating machines;
2. The development of algorithms for planned calculations based on a unified sys-
tem of information;
3. Dynamic modeling for developing the national economy; and
4. Mathematical work for constructing a unified, centralized national economic
plan, which would develop “the communist form of self-government of the produc-
tion units, the optimal composition of general governmental interests, every com-
pany, and every worker.”58

These quotations came from his earlier discussion of Lenin-Marxist rhet-

oric for economic planning. By wedding the cybernetic and Marxist-Lenin-
ist rhetoric of self-governing economies, Nemchinov sought to propose
“economic cybernetics” and its plausibly nonsocialist “dynamic models of
balancing capital investment” in the ideologically most acceptable light.59
CEMI—in Nemchinov and his superiors’ original vision—was set to become
a powerhouse intellectual engine for driving a cybernetic vision of the net-
worked national economy.
In late 1962, after receiving preliminary confirmation that CEMI would
be established, Nemchinov, then age sixty-eight, grew too sick to continue
his work and transferred the directorship of the Institute to the young acade-
mician Nikolai Fedorenko. Nemchinov died November 5, 1964, at the age of
seventy. Had Nemchniov not grown ill, it is likely he, not Nikolai Fedorenko,
would have emerged as Glushkov’s first and strongest ally in Moscow.

Fedorenko and Glushkov: A Partnership Pulled Apart

In the beginning, Nikolai Prokof’evich Fedorenko proved a valuable col-

league, confidante, and foil for establishing Glushkov’s OGAS Project (fig-
ure 4.14). In 1962 and 1963, both cyberneticists were appointed the first

9800.indb 139 6/2/16 3:05 PM

140  Chapter 4

directors of brand-new and prestigious academic institutes—Fedorenko’s

CEMI in Moscow and Glushkov’s Institute of Cybernetics (IK) in Kiev—that,
under their directorship and their shared vision of networking the national
economy, led the Soviet Academy of Sciences in economic cybernetics. At
the start in 1963, this dynamic duo of rising young academicians seemed
destined to follow parallel paths to greatness while salvaging the failing
Soviet economy along the way—at least according to Aleksei Kosygin, then
deputy chair of the Council of Ministers, who was supporting their initia-
tives at the same time that he was advancing the liberal economic reform.
All lights appeared green, and in 1964, the funds began to pour into these
institutions to shore up the alternative to the Kosygin-Liberman reforms.
The personnel at the two institutes multiplied exponentially. In a few
years, the Institute of Cybernetics’ staff numbers grew from dozens to over
two thousand, and the ranks at CEMI sprouted from its original fourteen
researchers in Akademgorodok to over one thousand researchers and staff
in Moscow. Most of those new employees were young researchers with bold
ambitions and a distaste for the culture of totalitarian control in the 1940s
and 1950s. Enthusiasm for decentralized economic reform met with central
flows of funding. In the late 1960s, after construction work was complete,
CEMI moved into a state-of-the-art, twenty-floor skyscraper in the desir-
able Cheremushki neighborhood in Moscow, and after a decade of transi-
tion in the 1960s, the Institute of Cybernetics occupied a well-equipped
campus along the scenic southwest edges of Kiev (figure 4.15). At least for
a moment in the heady transition of 1962 and 1963, the two institutes
appeared ready to remake the Soviet economy together.
One of the systemic sources of institutional volatility in the Soviet
knowledge base was the oversized influence that individual leaders, like
CEOs in modern Western culture, played in navigating and mobilizing
organizational pursuits. In this sense, institute directors, such as Nikolai
Fedorenko, appear entrepreneurial in the almost conventional sense of
organizational leaders who take risks, invest in them, and mitigate the con-
sequences of those risks by creative institutional problem solving. A year
after the founding of CEMI, Fedorenko reported to the Presidium of the
Academy of Sciences, USSR, that thanks to “the Institute work on the cre-
ation of methods of optimal planning … savings [in the sector of transpor-
tation] have already reached about half a billion rubles.”60
CEMI, in its early years, was not bound by the institutional logic of path
dependence. Compare, for example, the major research directives that
Fedorenko lists in his yearly reports between 1964 and 1969. In the first
annual report (1964), Fedorenko lists the following six research directives,

9800.indb 140 6/2/16 3:05 PM

Staging the OGAS, 1962 to 1969  141

Figure 4.14
Nikolai Fedorenko, date unknown.

all of which concerned the building of an OGAS-related wide-area infor-

mation network (note the second and fifth, in particular): (1) develop a
theory of optimal planning and management for a unified mathematical
model of national economy; (2) develop a unified system of economic
information; (3) standardize and algorithmize the planning and manage-
ment processes; (4) develop mathematical methods for solving economic
problems; (5) design and create a unified state network of computer centers;
and (6) derive a specialized planning and management system based on
mathematical methods and computer technology. Five years later, by 1969,
that number had been pared down to three concerned with optimizing
and modeling microeconomic problems. The network initiative had disap-
peared from its Moscow initiative.
In other words, by 1969, the year that the U.S. ARPANET went online,
CEMI was no longer actively pursuing any unified computer network proj-
ects. As a RAND analyst noted in 1971:

The most conspicuous feature of the latest version [of CEMI’s research directives] is
the absence of any reference to the unified state network of computer centers. Also
missing is the proposed system of economic information. The projects, representing

9800.indb 141 6/2/16 3:05 PM

142  Chapter 4

Figure 4.15
Central Economic Mathematical Institute in Moscow, with Mobius strip statue, 2008.

research on the methodology of economic analysis and organization of new opera-

tional systems, are replaced by work on economic projects, a much less innovative
and more conventional activity.

In the same years, Fedorenko’s CEMI had drifted from the OGAS Project
and grew to nearly forty times the size of Nemchinov’s original laboratory.
At the beginning of the 1960s, the average age of its full-time faculty was
about twenty-six years old; ten years later in 1973, when the institute sur-
passed one thousand employees, the average age tallied in at thirty-four.62
According to Gavrilets, a lifelong faculty member at CEMI, the institute
began as a lively and energetic place for critical and enthusiastic young
economic researchers.63 Although an increase of eight years in the aver-
age age of staff members over a decade probably reflects natural aging,
CEMI’s workforce was still relatively young and energetic and conditioned
to believe they had the support to do anything. As a result, CEMI was not
constrained by any formal agreements, as OGAS campaigners might have
sometimes wished, to pursue OGAS and its associated network projects.
It might be that CEMI’s eventual abandonment of the OGAS Project
contributed to the failure of the USSR to reform its economic situation.

9800.indb 142 6/2/16 3:05 PM

Staging the OGAS, 1962 to 1969  143

The shortcomings of these technocratic economic reforms were due both

to the complexity of the reforms as well as the more foundational ad hoc
complexity of the ministerial networks that were scrambling for funds in
the first place. CEMI chose to devote its funding to microeconomic math-
ematical modeling of the economy (not the national networking of the
economy) because its success as an institute depended on its iterative
navigation and securing of state-approved funding. Instead of commit-
ting to particular projects (as Glushkov’s Institute did, in part thanks to
his personal leadership) or requesting and receiving funding to conduct
basic, unspecified research (as was common in both Soviet and U.S. mili-
tary spheres), CEMI had to defend and justify tens of millions of rubles in
expenditures for specified civilian-political purposes. Fedorenko reasonably
found linear programming and modeling optimal microeconomic interac-
tions (with what he called the SOFE method) to be a more sustainable and
less politically fraught task than networking the economy.
Funding of all sorts was earmarked for certain purposes, dependent on
budgetary categories, constrained by values set in institutional history and
shaped by practice, influenced by industry best practices, marked by gift-
giver sources, and saturated in the politics of negotiation and expectation.64
As the German sociologist and philosopher George Simmel maintained in
his classic work on money, economic value is as much a matter for the phi-
losopher and sociologist who debate orders of evaluation (or the realm of
the study of value) as it is for the accountants, for whom the key interest is
the measuring of monetary value based not on value itself but on the like-
ness or exchangeability of value.65 Fedorenko worked out the research direc-
tives for his explicitly civilian research institute in a decentralized funding
environment where economic value was subject not to a flat marketplace
but a hierarchy of state interests. As the beneficiary of such interests, CEMI
was free to redirect its research directives (in this case, away from the OGAS
Project) and was constrained to justify those civilian research directives in
politically acceptable terms (in this case, toward microeconomic model-
ing). The net effect of the decentralized funding environment for civilian
projects, especially during the political freeze under Brezhnev, for the Soviet
network institutional landscape was to redirect research toward more politi-
cally conservative agendas.
This conclusion might appear backward at first because CEMI’s choice
to focus on microeconomic modeling arguably shares more with liberal
economic (or neoliberal) calculation of value and the OGAS Project appears
to be a relatively conservative attempt to use technology to reaffirm and
rationalize the (decentralized) hierarchical command economy structure.

9800.indb 143 6/2/16 3:05 PM

144  Chapter 4

However, given the disconnect between practice and principal, the OGAS
Project appears both more philosophically bold and practically far-fetched
of the two economic cybernetic approaches.
Another barrier—the same that Kitov encountered in his show trial—
was the wall between civilian and military economies, and it began to
strain the hopes for an economic network. In the spring of 1965, Fedo-
renko and Glushkov approached the Ministry of Defense to discuss the
possibility of joining military network initiatives with their own OGAS
dreams. Both Glushkov and Fedorenko’s institutes were developing techni-
cally compatible, top-down, large-scale computer networks projects—and
as Kitov had pointed out in 1959, the Soviet military already had several
in operation.66 The military networks were hierarchical and decentralized,
loosely designed after the U.S. SAGE computerized air defense system, the
first large-scale computerized command-and-control system in the world.
And so with Kitov’s Red Book show trial in mind, Fedorenko and Glushkov
met with Defense Minister Bagramyan to discuss the matter. After an hour
discussion in which Glushkov and Fedorenko did most of the talking, the
Minister of Defense Bagramyan replied, according to Fedorenko’s memoirs,
with the following:

You are good men, and you are doing right by concerning yourselves with the econ-
omy of the people’s money. But I cannot help you.… My friends, the state gives me
as much money as I ask for to build the technical basis [of the network]. As far as I
understand, they give you nothing. If I were to cooperate with you, they would give
money to neither me nor you, since there is the opinion that economics is a scab on
the healthy body of the governmental mechanism for planning and management.67

In Bagramyan’s notion that “economics is a scab on the healthy body”

of the Soviet state, we encounter a conflict of organizational self-interest.
The Soviet military, which was the single greatest benefactor of the Soviet
command economy, refused to cooperate with a civilian cybernetic proj-
ect because of the prevailing disdain for the very economic management
techniques that the cyberneticists were hoping to reform. This denial of a
request for cooperation is an example of the unregulated freedom that the
minister enjoyed when he acted in what he felt was his institution’s best
interest. This organizational dissonance repeatedly overwhelmed Glush-
kov’s and others’ attempts at systemwide collaborative reform.
After this encounter, it is unclear how far, if at all, Fedorenko pur-
sued funding or collaboration with Glushkov’s OGAS. As in the case of
Paul Baran at RAND, funding decisions clearly favored defense rationales.
The Minister of Defense was free not to cooperate with anyone because

9800.indb 144 6/2/16 3:05 PM

Staging the OGAS, 1962 to 1969  145

top-secret military missions enjoyed competitive advantages over other

projects. This also meant that funding approval depended not on the will of
top Party officials but rather on peer and lateral coalition building among
organizations that were both cooperating and also competing for limited
funding and influence in the Soviet state. This contradictory institutional
space, where entrepreneurs seek to leverage organizational dissonance,
exemplifies what I mean by heterarchy. Heterarchy describes the presence
of ambiguities that result from competing formal regimes of evaluation,
and entrepreneurs are those who trade on those ambiguities.68 As a case in
point, the Politburo claimed to oversee the goals of both the Ministry of
Defense and Glushkov’s and Fedorenko’s institutes, and yet the Minister
of Defense operated within heterarchical power structure that gave it no
reason to recognize the Politburo’s evaluation of the OGAS. To have done
so would have questioned the necessity of the Ministry of Defense’s own
access to massive funding from the Politburo. Ministries, free of any single
centralized operational logic that might be capable of legislating coopera-
tion top-down, were free to not cooperate. They were also free to shut out
peer-competitor institutions.
By the late 1960s, CEMI under Fedorenko’s leadership had abandoned
the OGAS and EGSVTs national network project to refocus efforts on the
microlevel linear modeling of Soviet factories and enterprises. Fedorenko,
a former chemist who was accustomed to microanalytic scales, claimed
that CEMI’s contributions to analyzing the national economy had better
chances when applied to smaller, more manageable local scales, which
his institute developed into the optimal mathematical planning method
known as SOFE (System of Optimal Functioning of the Economy). In his
memoirs, Fedorenko admits that the number of successful macrolevel eco-
nomic analyses that CEMI produced in three decades “could be counted
on one hand.” In contrast, in the tally of firm-level analyses or smaller,
Fedorenko counted hundreds of successes over several decades of work.69
A closer look at CEMI’s stepwise separation from Glushkov’s OGAS Proj-
ect in the 1960s sheds some light on the negotiated compromises and qual-
ities possessed by entrepreneurs like Fedorenko in the Soviet knowledge
base. CEMI, under Fedorenko, went onto pioneer microeconomic modeling
across the nation. In 1964, CEMI opened a branch in Tallinn, Estonia, and
in 1967, a branch in St. Petersburg. In the 1966 preparations for the celebra-
tion of the fifty-year anniversary of the Soviet regime, Fedorenko described
the EGSVTs (network) project in glowing if slightly scaled-down terms: “An
important direction of CEMI’s research is the development and creation of
a unified state network. This network should consist of three levels: a main

9800.indb 145 6/2/16 3:05 PM

146  Chapter 4

computational center, a few dozen prominent computational centers, and a

lower network. Such a structure will allow flexible information accounting
and both operational management of the industry according to territories
and the organization of planning accounts according to topics.”70
After 1967, CEMI internal documents stop mentioning any network
projects, whether OGAS or EGSVTs. What began as a small laboratory
devoted to a wide-ranging civilian-use network for the management of the
economy became, as a RAND report later called it, an “operational support
agency” for the Gosplan.71 Today CEMI is remembered for spearheading
optimal planning methods with computerized and mathematical models
in the Soviet socialist economy. As its website proclaims with silent hind-
sight on its early network ambitions, “When the Institute was founded in
1963, its main goal was to elaborate the theory of optimal management of
the economy, applying mathematical methods and the use of computers to
the task of practical planning.”72
Like most rivalries, the subsequent rivalry between these two peer insti-
tutes, CEMI and IK, developed out of, in Freud’s phrase, a narcissism of
petty differences. After having their original OGAS mission tabled, both
resorted to developing from the bottom up microlevel, factory-level eco-
nomic planning. Even today, CEMI continues to pursue enterprise-level
economic modeling, and IK continues to develop automated systems of
management (ASUs) for individual enterprises. Fedorenko reported hav-
ing improved hundreds of factory-level flow models every decade, and
Glushkov claimed to have established ASUs in Ukraine, St. Petersburg, and
beyond. Despite these successes, Glushkov in 1975 observed that humans
entering “half-truths” were hampering automated control systems so that
“we find ourselves somewhere between confusion and a search for scape-
goats.”73 The few ASUs that were implemented fell flat, as well, accord-
ing to the émigré mathematical economist Aron Katsenelinboigen, who
reported that ASUs had little to no effect and sometimes even negative
effects due to the expenses of installation. Managers, who were often older
and wary of being replaced, often lacked the capacity to become familiar
with, let alone master, the economic-mathematical methods that the ASU
required.74 As Glushkov later noted in Pravda, one automatic control sys-
tem was dismantled and sold because it “impartially pointed out manage-
ment’s blunders and omissions.”75 What began as an alliance in the early
1960s around a network became a rivalry after the 1970s when cybernetic
institutes disagreed over the relevance and proper role of the computer in
economic planning. As these sections illustrate, the tensions resulted not
from the roles of computing networks and information technology but

9800.indb 146 6/2/16 3:05 PM

Staging the OGAS, 1962 to 1969  147

rather a series of serious administrative, institutional, personal, political,

policy, and social problems.

Management Missteps: “Supervision” and the Separation of the OGAS

and the EGSVTs

In 1962, after Keldysh advised Glushkov to submit the OGAS proposal to

the heads of the Communist Party without the moneyless payment sys-
tem, Glushkov, backed by Fedorenko and others, submitted his original
OGAS proposal for a chain of reviews by a number of Soviet government
agencies. As a result, a commission was formed to review his proposal,
which received preliminary approval, and in 1963, it arrived at the desk of
the Party Central Committee and the Council of Ministers. At this point,
Glushkov, Fedorenko, the chair of the Central Statistical Administration V.
N. Starovsky, the first deputy minister of communication A. I. Sergeichuk,
the vice minister of finance, and others gathered together as a commission
to discuss and review the proposal and several thousand pages of associ-
ated materials. For months in 1963, the commission met and discussed
the details of Glushkov’s proposal, and each member tried to object to and
reject specific measures in it. Despite the proposal’s considerable political
support to this point, including review by the Politburo and the Central
Committee, the result was support for a technical computer network but
not economic reform. For a period of time, the shell of the OGAS Proj-
ect was approved for “finalization” at the hands of the Central Statistical
Administration, and the heart of the OGAS economic reform was post-
poned until future review.
Thus, a technical network project—the EGSVTs—was born, and the auto-
mated management of the OGAS was put on hold. The technical network
was deemed the Unified State Network of Computing Centers (EGSVTs, for
edinnay gosudarstvennaya set’ vyicheslitel’nikh tsentrov), and in response, the
committees issued a joint decree titled “On Improving the Supervision of
Work on the Introduction of Computer Technology and Automated Man-
agement Systems into the National Economy.”76
The presence of the word supervision in the decree title here is telling. The
government agreed to improve the supervision of the automated manage-
ment of the economy, not management itself, which the top Soviet leaders
recognized must be left to the distinctly not automated human bureau-
cracy of state employees and planners. In particular, the officials charged
with approving the OGAS stumbled over Glushkov’s distinction between
a system that would make executive commands, which they feared, and

9800.indb 147 6/2/16 3:05 PM

148  Chapter 4

a system that could command information about those commands—or

the economic metadata. When faced with the possibility of controlling
all economic information, the commission reviewers concluded, given
the already tumultuous economic supervision in the early 1960s, that a
national economic network could supervise, not directly manage, the com-
mand economy.
Similar to Kitov’s Red Book show trial, the official rationale for the initial
decision to strip the OGAS of any capacity to reform the planning pro-
cess itself came with a justification that did not quite match the action.
In this case, the Central Committee denied the automated management
portion of the OGAS proposal due to what they deemed (not without
contradiction) to be the inefficiency of rational management systems. In
practice, the Committee apparently denied the request out of a fear that
Glushkov’s OGAS would strip its own unsanctioned informal control over
economic power. Commission members who supported the OGAS also
worried that even with top-level support, midlevel administrators would
sabotage OGAS’s efforts to rationalize their management powers. The ini-
tial 1963 decision to postpone the capacity of OGAS to reform economic
planning took place as Khrushchev was falling out of power and limited
Kosygin-Liberman liberal economic reforms were being introduced. The
submarining of both reforms highlights the contradictions that faced the
commanding heights of the Soviet state. No matter how obvious it was that
the mismanagement of the command economy drove the state’s economic
woes, the state could approve no major reform without a sweeping revolu-
tion in how it managed itself.
Glushkov learned his lesson from the 1963 commission experience and
scaled back and reframed his work on networking the command economy
from direct management to indirect information supervision. Beginning in
1963, he publicly repeated that “the OGAS does not command the econ-
omy, rather it commands the flows of information about the state of the
economy,” although in theory and practice, Glushkov grasped the inherent
politics of recordkeeping.77 There are good reasons to doubt this position as
a political compromise. As Glushkov theorized elsewhere, (1) a strict divide
between data and metadata functions denies the basic cybernetic propo-
sition of feedback loops that ensure that metadata observation is never
influence-neutral, and (2) no organizational reform can ever be divorced
from its political implications. Whether for political protection or other-
wise, the OGAS team, not unlike other information omnivore projects,
sought to ease its critics’ concerns by asserting that it would traffic merely
in metadata.

9800.indb 148 6/2/16 3:05 PM

Staging the OGAS, 1962 to 1969  149

In subsequent OGAS preparatory proposals, Glushkov reframed the

barebones EGSVTs network not as a matter of direct economic manage-
ment but rather as a support for information management related to the
national economy. This was so even though, in practice, the network he
proposed also advanced data exchange and communication across the local
and national levels. In the 1960s, Glushkov and his team tried at least two
times to propose to Party leaders a technical EGSVTs network—first as an
all-nation network (in 1963) and later as a regional network in Ukraine
(in 1967).78 A metadata management view informed both proposals: “To
organize information flows on the national scale,” as Glushkov once put it,
“one needs to centralize interagency management of all information banks
and computer centers, not the management of the economy.”79 In reorient-
ing his claim from the politically entrenched national economy itself to the
supposedly neutral territory of information banks, technical networks, and
data clearinghouses, Glushkov adopted the abiding belief that was com-
mon among cybernetics and many digital technologist heirs in the neutral
politics of code. Nonetheless, he proclaimed his task to be “not only scien-
tific and technical, but also political,” espousing the recurrent and trouble-
some idea that the politics of computation and technology are somehow
more neutral than other politics.80
Instead of imagining a future communism arising out of exchanges
ordered by an automated network, Glushkov envisioned the revised OGAS
Project as the means whereby human planners might process accurate
information about the economy via a national computer network. The
Soviet computer network, like similar computing projects elsewhere in the
1970s, appeared foremost to be a “public utility” and a mass medium for
serving information over great distance.81 (Computers too were mass media
in the age of mass media.) This revised model proved durable politically in
part because it came with the added efficiency of promises of liberal pricing
reforms, while capitulating to the more pragmatic demands of reforming
an economic planning administration staffed with self-interested humans.
Moreover, the revised emphasis on having OGAS manage the supposedly
immaterial information about economic interactions (rather than com-
mand the actual economic planning) also proved a salient political hedge
for the defense of the project going forward. Although striking near the
heart of the state communist project’s goal to transform the material well-
being of every citizen, the OGAS defenders publicly defended their reform
ambitions as merely immaterial and informational, even as the design ana-
log and cybernetic philosophy quietly espoused the more fundamental
fact that every information reform is also an organizational and thus social

9800.indb 149 6/2/16 3:05 PM

150  Chapter 4

reform. This convenient rhetorical distinction holds in later developments

of the OGAS Project, including Glushkov’s emphasis on “paperless infor-
matics” as a kind of successor to cybernetics as a theoretical vocabulary for
the emerging socialist information society.


Despite the tensions outlined above, the initial 1964 decision to down-
grade the OGAS from a full-service technocratic economic reform to an
EGSVTs technical network was sensible from the point of view of rational
state administration. The Soviet state was in a period of political and eco-
nomic transition from Khrushchev to Brezhnev, so it was not yet ready
to implement an economic reform like the OGAS. Its restructuring of the
information infrastructure of the command economy was so global that
it risked becoming a fully interactive networked political economy that
was run by remote-access data exchange and communication. In contrast,
the Liberman-Kosygin reforms invoked the scalable introduction of new
accounting profit measures in select enterprises and factories that, as the
liberal economists stressed, would cost no more than the stroke of a pen. In
comparison, the OGAS Project was too big to begin.
So as Kosygin began to implement the profit measure reforms in 1965,
the OGAS proposal suffered serious delays and was passed over for institu-
tional review for “finalization” by the Central Statistical Administration,
which was directed by one of the most outspoken opponents of the OGAS
Project on the commission, Vladimir Nikonovich Starovsky. Starovsky had
written to the chair of the Council of Ministers, K. N. Rudnev, as early
as November 1963 that he could not support the OGAS proposal because
it conflicted with the Central Statistical Administration (CSA) mandate to
oversee statistical matters, noting “a basic unified state network, in the
opinion of the CSA, should be the extant network of machine stations and
factories” already under its supervision.82 Starovsky’s opposition would
adjust but never reverse. In retrospect, Glushkov singled out Starovsky’s
resistance: “later, when the fate of the OGAS was being decided, the leader
of the Central Statistical Administration spoke against the project so much
more furiously than anyone else that he did much to seal its sad fate.”83
Still, the Central Committee did not reject the proposal and mandated
that the CSA would be in charge of finalizing the project. Stuck between a
rock and a hard place, Starovsky chose to resist by other means: the CSA
submitted the OGAS proposal for finalization review by sending it off to

9800.indb 150 6/2/16 3:05 PM

Staging the OGAS, 1962 to 1969  151

its most remote regional departments in Archangelsk and Karakalpak in

Siberia, where it underwent several years of what OGAS supporters recall as
a series of interminable and often incoherent feasibility reviews and often
nonsensical dataflow testing. The specific missteps of the Siberian CSA
review—such as arbitrarily declaring that after accounting for overhead
hardware costs, calculating economic problems by computer would be on
average ten times more expensive than calculating the same problems by
hand—are symptomatic of the information organization problem that the
OGAS sought to resolve and rationalize. In command economies, the more
information involved in planning, often the more opaque or meaningless
that information becomes. (It was never clear why computing by machines
should be ten times more expensive per calculation than by hand, and yet
the number stood with the force of administrative fiat.) Starovsky was con-
cerned that the OGAS would wrest from the CSA its central task of gather-
ing statistics for managing the command economy, and so by introducing
and inventing dubious feasibility information about an already uncertain
OGAS Project, he effectively stalled the economic reform portion of the
proposal from making progress at the national level through the rest of the
The global-local character of Glushkov’s decentralized design was part of
the genius of the project and also illustrates how the OGAS Project continu-
ously threatened the economic bureaucracy that it was meant to reform
and serve. Glushkov’s decentralized design of rational management could
work only if it was implemented top-down with the support of a central-
ized administration, such as the CSA. But no centralized administration
could be found to support it because, in practice, centralized administra-
tions in the civilian sector benefited from not behaving like centralized
administrations. Here we can begin to see the political paradox that the
OGAS encountered—one that was manifest in the tensions between the
formal master plan and the informal practices of the Soviet system and also
in the life and work of one of the Soviet master mergers of theory and prac-
tice. Glushkov was aware that it was in the self-interest of institutions in
the Soviet knowledge base to resist the OGAS. Despite having unparalleled
insights into how official and shadow economies worked, Glushkov had no
other choice but to model the OGAS network after the formal command
economic model, not after economic behavior.
Part of that design choice is an intellectual consequence of the cyber-
netic instinct to analogize the social and technical into structurally simi-
lar information systems, such as the command economy and its national

9800.indb 151 6/2/16 3:05 PM

152  Chapter 4

network. But a great portion of the design choice to model the OGAS after
the formal command economy follows from political necessity. Consider
this contradiction of political practice. To implement a fully decentralized
reform to the economy, top political support needed to be secured to imple-
ment the reform systematically. The full decentralizing reform first had to
be implemented with centralized systematic approval of the top. To gain
the support of those top central authorities, the reform design had to con-
form to the publicly approved and ideologically acceptable principles of
the current economic organization, which means that the OGAS design
had to map onto the pyramid structure of economic planning in principle.
So far, there is no contradiction because the short story of the tumultuous
history of Soviet economic reforms is effectively one of top leaders who
variously attempt to reaffirm their own hierarchical control, no matter how
The contradiction lies in the practical need for the reform in the first
place. The need for decentralized economic reforms follows from the fact
that, as discussed, the command economy in practice never functioned in a
strictly centralized manner. OGAS supporters sought to transform the econ-
omy into a decentralized hierarchy, but the economy, whose leaders publicly
defended their positions in a centralized hierarchy, never behaved as a strict
hierarchy because those leaders and their supporting personnel benefitted
by the informal economy of favors and heterarchical connections. Many
of those in the economic bureaucracy resisted the OGAS because although
it purported to support the formal power structure that legitimated their
positions, it also threatened to strip their institutions of the thing that jus-
tified their existence—the need to manage the command economy in the
first place. The OGAS, if effective, would strip those positions of what made
them informally beneficial to hold—the potential for corruption and per-
sonal gain and power. The organizational dissonance coursing throughout
the command economy both motivated the reform and caused this initial
frustration. With no other choice but to appeal to the top, the OGAS Proj-
ect was stranded by the potential adopters of its decentralized design (the
CSA in the late 1960s and other institutional entanglements in the 1970s
and 1980s) because the project sought to resolve the conflicts of interest in
the command economy that kept its own bureaucracy from resembling in
practice the pyramid of political power that it had to appeal to.84
OGAS did not meet its end at the hands of stalled feasibility reports by
the Central Statistical Administration between 1964 and 1969, however.
During these years, Glushkov, among others, built considerable politi-
cal support for developing the technical network of the EGSVTs. The late

9800.indb 152 6/2/16 3:05 PM

Staging the OGAS, 1962 to 1969  153

1960s were a helpful preparatory period for building and securing political
alliances that a small group of cyberneticists and network entrepreneurs
attempted to form in the 1970s. This period was spent quietly and care-
fully working within the administrative heterarchy to secure political sup-
port. To a surprising degree, Glushkov succeeded in doing so at the upper
echelons of Soviet power. Two top-ranking powerbrokers offered relatively
unwavering support of the OGAS Project in the late 1960s. First was Aleksei
Kosygin, who was effectively second only to Brezhnev in civilian matters.
He was initially chair of the State Planning Committee (Gosplan) and first
deputy chair of the Council of Ministers under Nikita Khrushchev (1959–
1964) and rose under Brezhnev to become premier of the Soviet Union. As
already noted, when Kosygin’s initial profitability reforms in 1965 were met
with fierce resistance from the economic administration and effectively
stalled, Kosygin turned to the OGAS as the next best approach. Second was
Dmitry F. Ustinov, who was a prominent military leader and manager who,
just before helping ousting Khrushchev in 1964, served as first deputy pre-
mier with control over the civilian economy. In addition to being a career
member of the Central Committee beginning in 1952, Ustinov ruled as the
leading defense minister of the Soviet Union from 1976 to 1984.
Not long after the commission decided to postpone the OGAS in 1964,
Petro Shelest—the first secretary of the Ukrainian Communist Party—called
Glushkov to persuade him to cease promoting the OGAS and return to work
(as Fedorenko in Moscow had already begun to do) on local or microeco-
nomic systems. Gluskhov and his team complied with Shelest’s commands
and turned their attention back to developing local and regional comput-
ing centers that might later be connected by telephone and telegraph cables
(figure 4.16). Soon after, Dmitry Ustinov countermanded Shelest’s wishes,
at least for the military: Ustinov, who was on his way to becoming minis-
ter of defense (1976–1984), invited Glushkov’s team to build ASUs in test
military factories.85 Military support appears to have given the team the
administrative license to advance the cause of computing technology and
also to have ensured that their ASU work would not benefit or network the
civilian economy.
In the 1970s, several civilian factories received ASUs under direction of
the OGAS team. Most of these efforts were carried out from the bottom up,
although Glushkov and his team at the Institute of Cybernetics continued
to seek and occasionally secure top-level support in the 1970s only to see it
dissolve in committees convened by intermediary ministries. For example,
Glushkov, with the support of the director of the S. O. Petrovskiy television
plant, successfully developed in two years local control systems such as the

9800.indb 153 6/2/16 3:05 PM

154  Chapter 4

Figure 4.16
Inside an ASU: Machine Hall, State Institute of Computing Centers, unknown date.

L’viv System or Lviv MICS—an automated control system for streamlining

the industrial processes in the Elektron television factory in L’viv, Ukraine.
After completing the L’viv System, the team engineered a more complicated
Kuntsevo system for planning and managing the resources of the Kuntsevo
radio manufacturing plant in southwest Moscow.86 The Institute of Cyber-
netics also proposed an industrywide network of ASUs in the industry-rich
Donbass region of Ukraine (figures 4.17 and 4.18).
Not all installations went smoothly. One factory manager, Valentin
Zgursky, senior technologist at a manufacturing plant, admitted that “when
you brought the Universal Control Computer [a mainframe behind the
ASU] to our plan for mass-production,” Malinovsky recalls being told, “I
did everything possible to make sure it would never succeed!”87 Neverthe-
less, Zgursky eventually saw the value of the ASU and installed it (although
his admission may have been the exception in the long run). Bolstered by
some local successes on the edge of an empire in the late 1960s, Glushkov
also repeatedly reminded anyone who would listen about the work that
even a dozen or so local systems (ASUs) could do after they were connected
into a single national network.

9800.indb 154 6/2/16 3:05 PM

Staging the OGAS, 1962 to 1969  155

Figure 4.17
Diagram of an ASU network at the industry level, about 1969.

In the late 1960s, after these and other limited local successes, top lead-
ers began to heed some of Glushkov’s calls more carefully. Dmitry Ustinov
commanded the heads of the military ministries to heed Glushkov’s orders
while he continued work on the L’viv System. After securing Ustinov’s top
brass support, Glushkov claimed that as early as the late 1960s, the auto-
mated management systems in factories throughout the empire provided
the outline of what would become the OGAS: “it was planned from the
very beginning that the whole system would apply across all spheres at
once, so some rudiments of an all-state system were conceived”88 (figure
4.19). After this chapter’s discussion of the bold vision and rocky institu-
tional landscape that supported the OGAS Project, the following chapter
chronicles and comments on what happened when, in 1970, the Soviet
centralized command decided to review the OGAS proposal to decentralize
the economy by network in earnest.
In summary, in this chapter I have examined the OGAS design think-
ing that motivated Glushkov and his teams and the initial obstacles that

9800.indb 155 6/2/16 3:05 PM

156  Chapter 4

Figure 4.18
Map of the proposed ASU train industry in the Donbass, Ukraine, about 1969.

Figure 4.19
Viktor Glushkov giving a presentation on the ASU, about 1969.

9800.indb 156 6/2/16 3:05 PM

Staging the OGAS, 1962 to 1969  157

they encountered. Against the bold vision of a networked electronic social-

ist future, a tangle of historical episodes frustrated the realization of that
vision. This chapter has offered a look at the institutional landscape and
alliances that formed and then dissolved between Nikolai Fedorenko’s Cen-
tral Economic-Mathematical Institute and Glushkov’s Institute of Cyber-
netics, their heydays as the leaders of economic cybernetics and networked
cybernetic reform through the late 1960s, the informal work culture of the
Kiev-based cyberneticists in the 1960s, and early bureaucratic barriers that
slowed the advance of the OGAS Project in the Soviet military. Neither
the Ministry of Defense nor the liberal economists wanted to collaborate
and support the OGAS Project, perhaps because the country had endured
four turbulent years, from 1962 to 1966. In that time the Soviet Union
had agreed to pursue computer-aided economic reforms, come to the brink
of nuclear disaster in Cuba, forced out and replaced its general secretary,
founded and funded leading economic cybernetic institutes devoted to
building a national network plan, foregone approving the original pro-
posal to introduce liberal profit reforms, and continued to fund the leading
economic-mathematical research institute in Moscow as it reoriented itself
away from its original network resolution to focus instead on less risky local
optimization and modeling problems. Topsy-turvy institutional behavior
in civilian matters was the rule, not the exception.

9800.indb 157 6/2/16 3:05 PM

9800.indb 158 6/2/16 3:05 PM
5  The Undoing of the OGAS, 1970 to 1989


No single actor could either make or undo the OGAS (All-State Automated
System) Project. The hidden networks governing the Soviet state were far
too complex and heterarchical to have had any single cause. (Most multiac-
tor networks involve complexities that are impossible to express in linear
form.) This chapter briefly outlines and analyzes the slow struggle over the
political execution of the OGAS Project in the 1970s and its aftermath in
the 1980s. The prolonged struggle and decline at the hands of various forces
helps to reveal the complex heterarchical forces that governed the Soviet
state and attempted to carry out economic and technological reforms. The
commentary that follows speaks by analogy to modern observers who are
concerned with attempts to reform complex political economic systems
and also reflects on how attempts to create formal computer networks are
sometimes thwarted by hidden social networks.
In this chapter, I chart the institutional apex, plateauing, and decline of
the most ambitious attempt to provide the Soviet nation with its own form
of networked socialism. The chapter begins by rehearsing the 1970 Polit-
buro review of the OGAS proposal, the ministerial defiance and contingent
institutional interests that extinguished its approval at the last minute, and
the subsequent dozen years (1970–1982) of attempts by Glushkov and his
team to revitalize state and then public interest in a networked socialist
economy. The chapter then takes a detour through an unlikely case study
before becoming reflecting on the central theme the military-civilian divide
that separates hierarchical and heterarchical institutions in the Soviet
Union. This case study examines how militarized strategic thinking—in the
hands of one of the great Soviet chess masters—materialized into a stillborn
attempt to plan the nation’s political and economic strategies with early
Soviet computer chess.

9800.indb 159 6/2/16 3:05 PM

160  Chapter 5

Ministry Mutiny

The strategic move that eventually stalemated the OGAS Project did not
come from abroad. It came from within. By the summer of 1970, Glushkov,
Ustinov, and others had mobilized enough support for a fresh review of the
OGAS Project by the highest committees in the land. All signs, except one,
suggested that the timing for an economic network renaissance was finally
right. The Central Statistical Administration (CSA) could no longer delay its
“finalization” review process for the OGAS proposal, which formally ended
in 1966 but had lingered in approval limbo ever since. Simultaneously,
the successful evidence of the local foundations of the EGSVTs (Unified
State Network of Computing Centers) was gaining more and more support,
especially as Party leaders searched for an untried approach to economic
reform in the wake of the faltering Liberman reforms. By the time that
Viktor Glushkov and Nikolai Fedorenko’s partnership drifted into a rivalry
over the wisdom of economic reform by macronetwork (Glushkov’s OGAS)
or micromodeling (Fedorenko’s SOFE, or System of Optimal Functioning
of the Economy), the EGSVTs had become such a promising project that
established rivalries were reigniting over whose administration might best
oversee its development and command the funding streams that came with
it. By early in 1970, Vladimir Starovsky’s Central Statistical Administration
and Vasily Garbuzov’s Ministry of Finance began to jockey for position to
command the administration of the OGAS Project. These two powerful min-
istries began contending not just for the project but against one another in
an effort to limit the competitor from securing massive funding.1
The most vocal opponent to the OGAS proposal in 1970 was also the
man who officially had been charged with its care and finalization for
the previous seven years. Vladimir Starovsky, the head of the Central Sta-
tistical Administration, “harshly objected to the whole project,” Glushkov
recalled in the late 1960s—out of opposition not to the economic reform
but to the prospect that the Central Statistical Administration would have
to cede control over some element of the governance of his administrative
turf (economic statistics) to future OGAS directors. Starovsky rejected the
remote-access portion of Glushkov’s proposal (a precursor to “cloud com-
puting”). If realized, the OGAS was going to provide access to information
and processing power to any authenticated user anywhere on the network.
Even though the permission hierarchy for authenticated users presumably
could still reaffirm the strong hierarchical structure supporting his admin-
istration, Starovsky opposed what we now recognize as a cloud computing
provision as being politically “unnecessary” because the Central Statistical

9800.indb 160 6/2/16 3:05 PM

The Undoing of the OGAS, 1970 to 1989  161

Administration was “organized by the initiative of Lenin” and already

does everything that Lenin asked of it. Reversing Lenin’s original question,
“What is to be done?,” Starovsky concluded that, because of Lenin, “Noth-
ing needed to be done.”2
From 1964 to 1970, as the CSA and Ministry of Finance were butting
heads, another front of intellectual opposition arose against the OGAS Proj-
ect from its own closest allies for economic reform—liberal economists. In
1964, a pivotal year for reform, Liberman, Belkin, Birman, and others were
able to convince Kosygin that, in contrast with the nearly 20 billion rubles
that the OGAS was predicted to cost, the cost of liberal economic reform
would be “no more than the cost of the paper on which the resolution of
the Council of Ministers would be printed.”3 Glushkov was caught unpre-
pared for this counterattack, having already admitted to Kosygin that the
whole network project would be net profitable but would prove to be more
costly and complicated than the space and atomic programs combined.
Nonetheless, the OGAS reform had the strategic advantage of not abandon-
ing Marxist planning principles for liberal market ones and of promising to
pay for itself quickly (and Glushkov foresaw a reimbursement of 5 billion
rubles by the end of the next five-year plan).

The Day of Reckoning: October 1, 1970

Several factors led to the Politburo’s review of Glushkov’s OGAS proposal

on October 1, 1970, which was the closest that the Soviet Union ever
came to approving a national network of its own design. In the midst of a
larger space and technology race, the unexpected revelation that the ARPA-
NET—the first American civilian nationwide network—had gone online
one year before, on October 29, 1969, suddenly hastened the search by
top Party leaders for a viable local national network project. Knowing that
the ARPANET was worrying Party leadership, Glushkov approached A. P.
Kirilenko, then secretary of the Central Committee, to ask the committee
to revisit the ideas in the previous proposal. Kirilenko welcomed the idea
and asked Glushkov to “write down in detail what has to be done,” said A.
P. Kirilenko, “and we will create a commission.” Glushkov wrote in reply:
“The only thing I ask is not to create a commission. Commissions operate
on the principle of subtraction of brains, not summation, and they can
wreck any project.”4
Nevertheless, Party leaders insisted on creating a commission. Glushkov
declined to chair it, and so V. A. Kirillin, then chair of the State Committee
for Science and Technology, was appointed as chair with Glushkov as his

9800.indb 161 6/2/16 3:05 PM

162  Chapter 5

deputy. The oppressive Soviet interventions in Czechoslovakia sent a wave

of recentralization, or rather antidecentralization, criticism through the
state, and some Gosplan officials openly criticized EGSVTs proposals. The
Politburo also felt pressed to consider and approve meaningful reform proj-
ects for the drafting of the Twenty-sixth All-Party Congress and the starting
of the eighth five-year economic plan in 1971. As a result, the Politburo
twice reviewed and approved without change Glushkov’s OGAS wordings
for the draft portion of the Twenty-sixth Congress. A preliminary meeting
of the same review commission (which had dragged its feet since 1964)
concluded in 1970 that the full OGAS, including the economic manage-
ment part, should be approved for top-level review, although who would
steer it after it was approved remained strategically unresolved. In particu-
lar, it was left unclear whether further “finalization” by the Central Statisti-
cal Administration would be required.
This time only one person on the review commission did not sign onto
the newly revived OGAS proposal—the minister of finance, Vasily Garbu-
zov, who was the primary opponent to the CSA. Garbuzov refused to sign
because he did not want the OGAS to fall under the control of his com-
petitor institution, the Central Statistical Administration, whose director,
Starovsky, also temporarily withdrew his support for the same reason sev-
eral years earlier. Glushkov and his team deliberated over how to proceed.
He did not want to submit his proposal to the Politburo for review if it
lacked unanimous support, but he also knew that he could not resolve Gar-
buzov’s concerns. Thus hedging its bets and hoping that the U.S. ARPANET
would sway the Politburo into action, the commission (unofficially led by
Glushkov) submitted the proposal for review.
That fateful gathering took place in Stalin’s former office in the Krem-
lin. As Glushkov walked into the long, red-carpeted room, Kirillin, one of
Glushkov’s supporters in the Politburo, leaned over to whisper that some-
thing had happened but he did not know what. Before he could clarify,
Glushkov noticed that something was out of place: the seats of the two
most powerful men who should have been in the room were empty. Gen-
eral Secretary Brezhnev and his prime minister, Aleksei Kosygin, did not
see eye to eye about many things, but on the matter of network economic
reform in the fall of 1970, they appeared to be ready to make an uneasy
truce. As it happened, Secretary General Brezhnev, who was a technocrat
with an engineering background and was favorably inclined to sweeping
technocratic solutions (especially those that disadvantaged orthodox eco-
nomic planners), happened to be away for the day in Baku attending the
fiftieth anniversary of the Soviet rule in Azerbaijan. Glushkov might have

9800.indb 162 6/2/16 3:05 PM

The Undoing of the OGAS, 1970 to 1989  163

counted on Kosygin’s support, but he too was away, pressing hands among
the mourning crowds in Cairo at Gamal Abdel Nasser’s funeral, who had
died of a heart attack two days earlier. Both men—the first and second in
command, including the economic reformer who was most likely to lobby
for the OGAS Project—could not attend the fateful meeting because of cal-
endar contingencies.
Despite these key empty seats, the meeting began well enough. Without
Brezhnev and Kosygin in attendance, the meeting was conducted by the
Stalinist-era Mikhail Suslov, who was famous for resisting radical changes as
the “Chief Ideologue of the Communist Party” and a consummate behind-
the-scenes operator with seats on both the Secretariat and the Politburo.
Given this steely reputation, he began encouragingly by saying nothing
against the proposal. Glushkov was then invited to speak, which he did
briskly before responding to a series of questions to the apparent satisfac-
tion of all involved. This went on for less than half an hour, until several
higher-ups began to speak positively about the project. Baybakov, one of
Kosygin’s deputies, volunteered that if the Politiburo should make him
head of the State Planning Committee (Gosplan), he would eliminate or
merge three ministries so that staff could be found to support the OGAS
Project. In this deft maneuver, Baybakov managed to relay Kosygin’s enthu-
siasm and promote his own career. The Minister of Instrument Making,
Automated Equipment, and Control Systems (Minpribor), K. N. Rudnev,
had extolled the virtues of information technology in economic planning
in 1963, signed the document, and commented off the record that the
timing might be bad.5 A chorus of voices countered these hesitations with
unambiguous support of OGAS.
Just as it seemed that the committee might be nearing consensus
approval, the minister of finance, Vasily Garbuzov, stood up. According to

[Garbuzov] entered the stage and addressed Mazurov, Kosygin’s first assistant. He said
that, well, he went to Minsk as directed, to examine the poultry farms. At the so-and-
so farm, the workers designed a computing machine on their own. I laughed out loud.
He shook a finger at me and said, “You, Glushkov, shouldn’t laugh. We are discussing
a serious issue.” However, Suslov interrupted him: “Comrade Garbuzov, you are not
the chairman here, and it’s not up to you to control the proceedings of a Politburo
hearing.” He shrugged and self-confidently continued, “The machine can perform
three programs—turns on music when the hen lays an egg, turns lights on and off, and
so on. This increased egg production at the farm.” So he suggested that first we should
implement these machines at all the poultry farms in the Soviet Union and only then
could we even begin thinking about silly projects like a nation-wide system.6

9800.indb 163 6/2/16 3:05 PM

164  Chapter 5

At that point, Garbuzov, who served as minister of finance for another fifteen
years until his death in 1985, made a counterproposal. The OGAS should
be released from the control of the Central Statistical Administration and
put under the direction of a new institute that should develop (as the com-
mission had insisted back in 1963 as no more than the EGSVTs barebones
technical network) computers with lights that flash on and off. “Everything
related to economics and the elaboration of mathematical models for the
OGAS, etc.,” Glushkov recalled, “was wiped off.”7 From a technical perspec-
tive, Garbuzov argued, the EGSVTs approach made political common sense.
A technical network would avoid the minefield of economics, politics, and
ideology without foreclosing the possibility of introducing relevant eco-
nomic programming into that network in the future. This technical vision,
Garbuzov argued, was the most risk-averse way forward.
Behind the veneer of Garbuzov’s technical pragmatism lay a more self-
interested motivation for this counterproposal. Having not been able to
secure the OGAS for his own ministry, he preconditioned his technically
reasonable counterproposal on the fact that a new institute should be devel-
oped to oversee the OGAS. If his ministry could not have the OGAS, then
no other existing administrative entity should have it, he reasoned. After
all, by what other way can a minister reduce the bureaucracy except by cre-
ating a new bureaucratic body to do so? By so specifying, Garbuzov sought
to streamline the network development and submarine the chances that
this competitor organization, the Central Statistical Administration, had of
securing the massive funding streams and political gravity associated with
commanding the management and automation of the command economy.
As the Politburo discussion ensued, the consensus slowly shifted from
Glushkov’s OGAS in favor of Garbuzov’s EGSVTs counteroffer. At last Suslov
intervened, concluding the discussion with executive authority: “Com-
rades, perhaps we are committing a mistake by not adopting the project
fully, but it is such a revolutionary improvement that it will be hard for us
to realize right now. Let us do it that way, and we will see later how to pro-
ceed.” Suslov then asked what Glushkov thought, to which he responded
pointedly, “Mikhail Andrevich, I can only say one thing: if we do not do
[the full OGAS] now, then in the second half of the 1970s the Soviet econ-
omy will encounter such difficulties that we will have to return to this
question regardless.”8
Intrigue and unconfirmed speculation abound about how Garbuzov’s
Ministry of Finance managed to turn the Politburo against the OGAS
that day. Prime Minister Kosygin, who probably would have pressed for
a consensus in favor of the full OGAS, may even have chosen to attend

9800.indb 164 6/2/16 3:05 PM

The Undoing of the OGAS, 1970 to 1989  165

the Nasser funeral to avoid having to cast a negative Politburo vote on

the OGAS decision. Two years after the decision, in 1972, Glushkov heard
rumors about the apparent backstory behind Garbuzov’s counterproposal.
Before the October 1 Politburo gathering, Garbuzov purportedly sought a
private meeting with Prime Minister Kosygin to convince him that if the
CSA were allowed to govern the OGAS national project, the CSA would
grow so powerful that it could wrest control over economic matters from
Kosygin himself and the Council of Ministers, ceding it back to the Central
If the OGAS was approved, the minister of finance argued to Kosygin,
the Central Statistical Administration would surpass even Kosygin in eco-
nomic power. Garbuzov almost certainly did not make this warning out of
good will or concern for Kosygin’s position. His ministry had done the most
to undermine Kosygin’s political reforms during the prior five years. Since
1965, the Ministry of Finance had informally refused to implement the
Kosygin-Liberman reforms, thus encouraging discrediting criticism of the
reforms before they could take full effect. The winning argument appeared
to be a contradiction: Garbuzov contended that if Kosygin did not act to
preserve the status quo, Garbuzov’s competitor would strip Kosygin of the
power to make economic reforms. Faced with that option and ceding the
OGAS Project to the Ministry of Finance, Kosygin appeared stuck, although
whether Kosygin actually believed Rudnev’s argument does not matter
compared to the result. From late 1970 until his retirement in 1980, Kosy-
gin never moved to unmire the OGAS Project administratively.
Such were the moves and countermoves that were at work behind the
administrative end game of the OGAS Project in the Politburo. A more gener-
ous reading upholds the possibility that Kosygin, the great liberal economic
reformer, did not wish to yield to Garbuzov but nonetheless felt compelled
to do so because a disgruntled Garbuzov and his ministry might sabotage
any of Kosygin’s future attempts to make economic reforms, whether or
not the OGAS Project was governed by an independent administration.
For Kosygin, the decision to neglect the OGAS could have been the best
way forward in a lose-lose situation of mutually assured ministry mutiny
between the CSA and the Ministry of Finance—short of risking his own
power to make economic reforms. Because of the consequent ambiguities,
every administrator had to engage in a form of entrepreneurial negotia-
tion among their private plans, their competitors’ plans, and the state plan.
This tangled heterarchy at the top of the heap led every administrator with
a stake in the decision into a competition with his neighbors. Historical
contingency played a role, as well: perhaps there was to be no OGAS that

9800.indb 165 6/2/16 3:05 PM

166  Chapter 5

day simply because two seats were vacated by a general secretary who had
to attend a planned celebration and a prime minister who had to attend a
funeral. At least Kosygin could wash his hands of having to make any top-
level decision to advance the OGAS.
The OGAS Project was neither fully rejected nor approved. Instead, the
Twenty-fourth Party Congress in April 1971 agreed with the Politburo deci-
sion that the ninth five-year plan (1971–1975) would establish some skel-
etal semblance of the OGAS, including 1,600 ASUs (automated systems of
management); expand computer production by 2.6 times; and establish
a technical network, the EGSVTs, across the nation. The EGSVTs, in this
iteration, were to connect all higher-level branches and departments in the
planning administration, develop regional networks, and connect and con-
solidate the regional networks to the higher-level network. The proposed
details in 1971 were scaled back closer to the initial 1963 EGSTVs proposal
levels—with twenty to thirty regional centers and the piecemeal incorpora-
tion of the national economy lurking in the background.

The OGAS Project in Repose

Having secured partial approval for the second time in a decade at the hands
of a top-ranking commission but being no closer to his goal, Glushkov sol-
diered on in his commitment to introduce some kind of technocratic eco-
nomic reform. Despite the authorities (the three words that he used to title
his memoirs), Glushkov and his team installed ASUs (automated systems
of management) in local factories with the hope of one day connecting
them. Between 1970 and 1977, Glushkov and his team offered up a vari-
ety of decentralized network designs, although these proposals never sat-
isfied a wide range of relevant parties.10 A Ukrainian computing pioneer,
Boris Malinovsky—who for his technical and historical achievements should
be remembered as the dean of Soviet computing memory—claimed that
“Glushkov’s monumental efforts constantly ran into a wall of indifference,
misunderstanding, and at times, animosity in the top echelons of the com-
mand-administrative system.” According to Malinovsky, the Soviet higher-
ups who never publicly criticized Glushkov were Prime Minister Kosygin
and Defense Minister Dmitry Ustinov, although the proposal also elicited
resistance from lower-level figures.11 Nonetheless, after years of politicking
on behalf of the OGAS Project, Glushkov convinced the CSA to reinstate the
word OGAS into the 1976 report on its “Main Directions”—and breathed
new life into the core idea of a Soviet industrialist network that united auto-
mated control systems across national economic branches.

9800.indb 166 6/2/16 3:05 PM

The Undoing of the OGAS, 1970 to 1989  167

A year later, in 1977, the state decided to declassify the OGAS Project,
meaning that the OGAS was no longer a state secret. This decision reflected
the project’s declining strategic significance to the state as well as a shift
in Glushkov’s long-term campaigning. Before 1977, promoters accepted the
ban on public discussion in part because it meant that the secret project was
vital to the highest political echelons. But this secret classification also served
its opponents in the state because public circulation and promotion of the
OGAS could have curried public favor for what could prove to be a career-
threatening reform. After the lifting of the top-secret clearance, however,
this could change, and Glushkov successfully petitioned Pravda newspaper
editors to begin a campaign to promote the network project with his article
titled “The Matter of the Whole Country” in 1980 (although Malinovsky
notes in the English translation of the dual-language Store Eternally, without
clarification, that the published version of the title was actually “For the
Whole State”).12 The article’s publication in Pravda implies a mixed public
relations victory because appearing in the nation’s flagship newspaper meant
that its editorial board, the Central Committee itself, had deemed the project
to be worthy of public discussion and not one of its prized state secrets. (The
conclusion to this chapter returns to the issue of public discussion.)
With a sizeable audience for the first time, the OGAS Project diversi-
fied quickly into a number of complex possibilities in the hands of leading
academics such as Glushkov, V. A. Myasnikov, Yu. A. Mikheev, and others.
Under their leadership and assignment to build a technical network that
connected local factory control systems, a number of associated subproj-
ects arose, including the ACPR (the automatic system of planning accounts,
or avtomatizirovannaya sistema planovyikh raschetov), ASGS (the automatic
system of state statistics, or avtomatizirovannaya sistema gosstatistiki), the
ASUNTP (or automatic system of management of scientific-technical
progress, or avtomatizirovaanaya sistema upravlenia nauchno-tekhnicheskim
progressom), and the ASUMTS (automatic system of management of mate-
rial-technical supply, or automatizirovannaya sistema upravleniya material’no-
tekhnicheskogo snabzheniya).13 The subsequent multiplication of associated
ASU systems and subsystems in the late 1970s and early 1980s attests to
two underlying trends—first, a general academic (public) interest in the
OGAS Project across planning, statistics, science-technological revolution,
and supply institutions; and second, a splintering or at least division of that
overlapping interest into subsystems according to preexisting complex rela-
tions between branch, regional, and national economic planning interests.
The movement to “ASUify” the nation in the 1970s never met with con-
siderable success. Given that the introduction of an ASU to a factory or

9800.indb 167 6/2/16 3:05 PM

168  Chapter 5

enterprise costs on average about 800 thousand rubles (or roughly just over
$1 million U.S. in the 1970s or over $4 million U.S. in 2016), ASUs were
introduced slowly and steadily in the late Soviet Union. According to one
account, as few as twenty-nine ASUs were introduced between 1971 and
1975, thirty-two between 1976 and 1980, and thirty-four between 1981
and 1985. Another report holds that from 1971 to 1975, the number of
ASUs grew almost sevenfold, although, even if the OGAS Project were sud-
denly approved, they could not easily be unified.14 Other accounts mention
even higher numbers, including one that claims that between 1966 and
1984, approximately 6,900 ASUs of different configurations were estab-
lished throughout the USSR.15 This vast discrepancy underscores the point
that whatever systems were developed under a sweeping state mandate to
advance information technology throughout the country, they were done
so without the benefit of any organized coordination from the state. The
lack of coordination hurt the effectiveness of the OGAS Project. Official
statistics determined that the computer technology that was in place ful-
filled no more than a sixth of its projected capacity in the affairs of local
economic management.16 The introduction of more computing processing
power in the form of third-generation computers adopted from abroad sig-
nificantly altered these modest growth trends in the managing of the com-
mand economy. The effort to network local enterprises and factories was
met with resistance from workers and managers. There was brooding fac-
tory floor–level discontent with the local factory computer control install-
ments, which were the local nodes that someday might be connected to
form the EGSVTs and OGAS.17 The workers did not feel empowered by their
access to the circuitry of the state’s master plan because the master plan
exercised managerial oversight over only local factories. As in early comput-
ing industries elsewhere, the simultaneous development of different core
computer systems in different systems led to protracted interoperability
problems and technical delays.
The state secrecy that characterized the OGAS Project from 1959 to 1977
facilitated a kind of boundless technocratic imagination about the possibili-
ties of networked computing that was not tempered by the humbling reve-
lations of practical experience. Although the early developmental period of
the OGAS saw profound accomplishments (including the launching of sat-
ellites and astronauts in space, the harnessing of the atom, and the advance
of problem-solving machines), when those technological innovations were
applied to everyday routine operations and tasks (such as sending and
receiving economic information across the command economy and devel-
oping automated programs for deciding what to do with that information),

9800.indb 168 6/2/16 3:05 PM

The Undoing of the OGAS, 1970 to 1989  169

the experience of networked computing in the late Soviet period revealed

just how modest the accomplishments of most science and technology are
day to day. This widening breach between the grandiose intellectual pos-
sibility and the modest applied practicality was central to Glushkov’s local-
global approach to practical universals, but it had the unwanted effect of
dampening public and institutional enthusiasm for the unmet expectations
of the OGAS Project. Accusations flew among stalwart Communists, who
blamed the internal divisiveness and infighting among the top levels of
government on the interventions of skilled enemies, especially American
capitalists. However overgenerous to enemies’ prowess for subterfuge this
may be, one sympathizes with their frustration while doubting the utility
of such countercounter measures.
In their discursive move and countermove, the public debates about
networks in the 1970s are not exceptional for the period and place. Just
as Kitov, Lyapunov, and Sobolev had done in their initial article by claim-
ing that anticyberneticist Soviet philosophers had fallen victim to the
machinations of a subtle pro-American disinformation program, Glushkov
occasionally partook in that classic cold war move of blaming the cunning
enemy for one’s internal problems. Glushkov, for example, once blamed an
unnecessary political battle in the 1972 All-Union Conference on a “dis-
information campaign skillfully organized by the American secret service,
which was directed against the improvement of our economics.” No matter
how fueled by the fumes of international conspiracy, such claims appeared
to work at home. Once, Glushkov reports, he was able to soften the blow
of an internal attack on his local automated system of management (ASU)
work by asking the Soviet scientific adviser in Washington, D.C., to issue a
report on how the competitors to his proposed computer were becoming
less popular in the United States. The report was read widely in the Polit-
buro and had its intended effect, leaving Glushkov’s project on the table
and scuttling his competitor’s.18 The positive corollary abounds in practi-
tioner memoirs, where a colleague compliments an associate by attributing
retroactively visible similarities between friend and foe to the friend. For
example, “as a thinker, V. Glushkov distinguished himself by the scale and
the depth of his works,” notes the president of the Ukrainian Academy of
Sciences, Borys Paton (whom Glushkov served as vice president from 1962
until his death), and he continued that “he predicted many things that
appeared in the Western information society much later.”19 While traveling
abroad, Glushkov once declined a lucrative salary offer from IBM, which
also stands as a badge of honor. In both, rivalries imprinted images of per-
sonal hopes and fears onto the faces of doppelganger foes.

9800.indb 169 6/2/16 3:05 PM

170  Chapter 5

Few among the technocratic optimists or the disappointed practitio-

ners were prepared to make the more general observation that the OGAS
experience is not unusual in how its bold technocratic inventions and pro-
nouncements were followed by plateaus of technological innovation that
swept through the long story of Soviet history of technology and science.
The real story about Soviet computing networks has far less to do with the
technology itself than with the institutional, political, economic, and social
networks that made up the knowledge base and innovation infrastructure
in a country and culture.

Bureaucratic Barriers

Glushkov was clear about the sources of the frustration to his life work: cun-
ning enemies were not infiltrating his life work from outside the nation, but
cunning competitors from within were doing so. After the Central Com-
mittee’s partial rejection of the OGAS Project in 1971, rumors circulated
that his local enemies were conspiring against him. In 1972, the pilot of a
plane that Glushkov was flying in had to make an emergency landing and
discovered that the fuel had been tampered with. It was rarely cloaks and
daggers for prominent Soviet mathematicians, however. The most common
obstacle was the pragmatic apathy that prevailed against his ideas for tech-
nological reform. In response to the proposal for an electronic office, for
example, a commentator expressed doubt: “if it takes a month and a half to
act on a letter to a Ministry, no automatic letter opener is going to change
anything.”20 In his memoirs, he calculated the malaise that characterized
his meetings with government officials with a characteristic precision:
“Unfortunately, my organizational efficiency coefficient … did not exceed
four percent. What does that mean? It means that in order for a problem
to even be considered by the government, I had to speak with twenty-five
officials.”21 An inefficient bureaucracy was both the obstacle to as well as
the target of his technocratic reforms. In 1972, he illustrated this with an
eye-catching statistic: according to his estimates, at 1 million operations per
man with an adding machine, it would take “10 billion persons” “to solve
all of today’s management problems.” The same operational burden could
be handled by men and women at 25,000 to 30,000 Minsk-32 computers
(at 30,000 operations per second), and even that number would quickly
decrease as processing power continued to increase.22
Antibureaucratic sentiment is not uncommon among highly skilled
technical workers and even among other bureaucrats. In Glushkov (whom
Hoffman once described as “probably the most forceful Soviet advocate and

9800.indb 170 6/2/16 3:05 PM

The Undoing of the OGAS, 1970 to 1989  171

the bluntest critic of computerized communication in the USSR”), it took

a particularly acute form. Again, recall that after the Central Committee
heard that the ARPANET had gone online and effectively granted Glush-
kov a blank check, the secretary of the Central Committee asked Glushkov
to “write down in detail what has to be done,” said A. P. Kirilenko, “and
we will create a commission.” Glushkov’s response is reminiscent of Baran
before he withdrew his network project from consideration by the U.S. mil-
itary. Both insisted that, whatever else happened, their network projects
not be handed over to the available administrative entities.”23
In over a dozen interviews that I conducted with scientists who were
associated or familiar with the OGAS Project, they unanimously complained
that bureaucratic infighting was the primary obstacle to their project. There
was something dreadfully wrong with the bureaucratic administration of
the national economy and the handling of the OGAS Project. But Glush-
kov’s critique looks beyond the bureaucrats themselves. Given a Weberian
understanding of bureaucrats as depoliticized professionals, many people
who held positions in the command economy and state were not rational
bureaucrats at all. They affected an “iron cage” of bureaucratic petrification
when convenient and waged war with other local deities. The problem does
not belong to all modern bureaucracy. Some bureaucracies do not result in
this kind of incessant, internecine Hellenistic competition among the gods.
Some administrations, including Soviet military ones, have successfully
managed to fund, develop, and launch megaprojects, and most large-scale
modern institutions are administered by functional bureaucracies.24 Where
lay the difference?
Glushkov believed that a successful bureaucratic system could be
reformed and improved with information technological upgrades, but only
with commensurate social and economic reforms. In other words, tech-
nical reforms to administrative systems without behavioral changes were
condemned to fall into a kind of double-bind: no minister could manage a
complex economy by paper, and yet no one with control over the papers
at hand would agree to switch to a “paperless” virtual economy, which
Glushkov had championed in the 1970s. Because the “chief content” of the
computing revolution was no less than a cybernetic fusion of information-
processing people and their machinery (in other words, “the appearance of
an essentially new man-machine technology for processing information”),
the success of any technical system reform would depend on social and
organizational changes: “Since the circular flow of information is the basis
for the functioning of any organization, [the information revolution] must
be viewed primarily as a revolution in organization and management.”25

9800.indb 171 6/2/16 3:05 PM

172  Chapter 5

For its main theorist, the OGAS Project could not meaningfully upgrade
to the command economy technologically without also simultaneously
reforming the organization and management of economic information.
Unfortunately, this separation of reforms is what the Central Committee
had repeatedly requested when it insisted that Glushkov begin with the
technical computer network EGSVTs before developing the automated sys-
tem of economy management that was central to the OGAS Project. He
frequently warned that without commensurate structural and behavioral
transformations of the economy, the introduction of information technol-
ogies would slow economic growth:26

The conservatism of the traditional technology for processing planning and man-
agement information leads to the intensification of “disorganized complexity” in
the national economy and erects informational-organizational barriers to planned
economic growth.… The problem, of course, is not just in the technology of orga-
nizational management. The economic mechanism plays a large (indeed a primary)
role here.… However, it is important to emphasize that economic mechanisms (es-
pecially under socialist conditions) do not work by themselves in isolation from the
organizational management system.27

In other words, perhaps the most direct cause for the failure of the OGAS
to develop, according to Glushkov, was rooted in the same motivation that
drove him to develop the OGAS in the first place—the observation that
the effects of the modern information science and technological revolution
cannot be separated from the social, economic, and organizational condi-
tions that shape them. The lot of networked computing cannot be under-
stood without the networks of institutions that first attempted to usher
technological networks into being.
This approach identifies at least two complementary organizational
barriers to the success of any attempt to systematic reform—centralized
self-interest and the decentralized status quo in Soviet society. First, had
networked computing been integrated into the fiber of Soviet society
(which it was not), it would have compelled broad-based systematic social
changes that could not have easily been isolated (as they usually were) to
the military industries.28 Because military interests maintained a strong self-
interest in preserving military power (and not social or economic progress),
these same organizations also actively resisted encouraging the develop-
ment or sharing the benefits of networked computing technologies outside
of narrow military applications. It was clearly in the military’s self-interest
to maintain centralized control over networked computing innovations.
Second, at the same time, the decentralized network of competing inter-
ests that governed nonmilitary administration also ensured that attempts

9800.indb 172 6/2/16 3:05 PM

The Undoing of the OGAS, 1970 to 1989  173

to introduce networked computing into social and economic planning

would break against well-organized centralized resistance from the military
and broad-based and haphazard resistance from anyone in a position to
benefit from the status quo. Because the OGAS threatened to reorganize
the social and economic spheres of life into the kind of rational planned
system that the command economy imagined itself to be in principle, it
threatened the very practice of Soviet economic life: networked comput-
ing, in Hoffmann’s analysis, “creates more choice and accountability and
threatens firmly established formal and informal bases of power through-
out the entrenched bureaucracies.”29 These two threads of analysis met in
the friction between a formal centralized hierarchy and an informal, decen-
tralized heterarchy. Both the military powers and the decentralized network
proposed by Glushkov were clearly hierarchical in operation. But the actual
workings of Soviet economic and social power were neither hierarchical nor
market. They were heterarchical, dynamic, and continuously reconstituted
in the interwoven political networks of social relations in the economic
bureaucracy facilitated by the Communist Party.
When asked why he thought that the OGAS did not take, Glushkov
responded with a comment that distinguishes military (space and atomic)
programs from the civilian administration:

S. P. Korolev (“the chief designer” of the Soviet space program) and I. V. Kurchatov
(the father of the Soviet atomic bomb) had a guardian on their side in the Politburo,
and they could approach him and immediately resolve any question. Our trouble
was that we had no one, and our questions were even more complicated because
they involved politics and any mistake could have tragic consequences. For that
reason, a connection with any of the members of the Politburo was that much more

Aleksandr Stavchikov, historical secretary of the Central Economic-

Mathematical Institute (CEMI), also commented on why the EGSVTs did
not develop successfully. According to his unpublished notes and personal
interviews, Stavchikov retroactively faults “the romanticism” of the insti-
tute for the “globality” of its early network designs, observing in hindsight
how Glushkov, Fedorenko, and others agreed early in the 1960s that any
attempt to plan the national economy in its entirety would have to be done
at the national level: “Certainly, an attempt to plan the national economy
of such a huge country on the foundation of one hugely proportioned
economic-mathematical model,” Stavchikov admits, “would be doomed to
failure from the start.”31
As for why the network was designed hierarchically, Stavchikov inti-
mated that the cyberneticists had no better choice. The reasoning for the

9800.indb 173 6/2/16 3:05 PM

174  Chapter 5

hierarchy—and not, say, a fully distributed design or even an unevenly

decentralized or heterarchical model—was a matter of reading the writing
already on the wall: Nemchinov and Fedorenko decided to “build the coun-
try’s unified net hierarchically—just as the economy was planned in those
days.”32 (The institutional histories of CEMI and the Institute of Cybernet-
ics conspicuously leave out the names of Glushkov and Fedorenko, respec-
tively, although evidence of the early alliance abounds in personal memoirs
and interviews.) Justified by a grand cybernetic analogy between the formal
design of the command economy and the formal design of the computer
network, Stavchikov reasons, any other network design would have been
politically unviable in a formally hierarchical command economy. The
network visionaries had no choice but to design a computer network that
matched a system that did not exist except, like the networks, on paper free
from the informal competitive practices of administrative-economic reality.
Design logics can be compelling—too compelling at times. The cyber-
netic analog between hierarchical economy and network also fit the politi-
cal values of the period. Fedorenko and Glushkov felt they had no other
choice: they had to align their technical national architecture with the
political system architecture. They also appear to have wanted to do so.
All evidence suggests that these leading cyberneticist entrepreneurs were
committed believers and practicing promoters of the official socialist ratio-
nales of the command economy, which also made them reformist critics
of the irrational status quo. For these network entrepreneurs—Fedorenko,
Glushkov, Kharkevich, and Kitov—the heterarchical competition at every
administrative level was the signal problem that was in need of a sociotech-
nical fix. According to CEMI Secretary Stavchikov, “In this, [Nemchinov
and Fedorenko] planned to use extant economic-mathematical methods,
allowing [them] to guarantee mutual conformity, the very best interdepen-
dence of the numerous units of the hierarchy downward and horizontally—
between the units of one level, as well as to develop new units.”33 In other
words, according to Stavchikov, these economic cyberneticists decided to
model the structure of the network after the structure of the socialist econ-
omy, in essence invoking the well-established trope of cybernetic thought
that technical systems share common information structures with social
systems—including mind-computer, body-machine, and society-media sys-
tems. The hierarchical form came as cybernetic analogic impulses such that
the decentralized network proposal was designed, according to Stavchikov,
to “guarantee mutual conformity” and “interdependence” between the
formal Soviet economic hierarchy and its technical network.34 The cyber-
netic instinct to design the OGAS after a nervous system for the national

9800.indb 174 6/2/16 3:05 PM

The Undoing of the OGAS, 1970 to 1989  175

economic body and not the nation as a brain follows from this felt obliga-
tion to “mutual conformity” between the national economic and network
The choice of the national hierarchy as the basis for their network design
was both the industry standard and necessary for those who were looking
to streamline a command economy that was both hierarchical on paper
and heterarchical in practice. The contradiction that was central to the
breakdown of the Soviet economic-administrative system lies between the
formal hierarchical design of that state and its own informal heterarchi-
cal networks of management as practiced by those who administered the
state. The endgame of the OGAS Project was found not by strategic prob-
lem solvers who were seeking to solve or finish the game but by those who
were seeking to extend perpetually their turn at the table of administrative
This is a distinct argument that is separate from the standard historical
accounts of the collapse of the Soviet network projects and sociological
accounts of the collapse of the Soviet Union itself. Most accounts posit that
the basic problems were one of a rigid, top-down hierarchical state. Noted
scholars since the 1990s have argued that the Soviet state and command
economy were fundamentally incompatible with the emergent, flexible
information networks.35 I believe that I have shown why that is wrong and
why it misses the greater problem. Instead of a fundamental incompatibil-
ity between vertical states and horizontal networks, the heterarchical ambi-
guities of Soviet administrative networks reveal too much, not too little,
flexibility in its capacity to generate organizational dissonance crisscrossing
and overlaying economic hierarchical structures with lateral conflicts of
private interest. The Soviet state was too familiar with the unpredictable
dynamism of competing informal networks (the same kinds of networks
celebrated by Internet commentators in the 1990s) to be able to carry out
systematic reform and infrastructure upgrade to bring the Soviet state into
the current network information age.

The Red King’s Book, or Botvinnik and the Soviet Case of Computer Chess

If war, in Carl von Clausewitz’s famous phrase, is a continuation of politics

by other means, then perhaps the most visible continuation of cold war
politics by means of a game is chess (second to Go, the world’s most popu-
lar war game). This classic thinking man’s game is synecdoche for cold war
confrontation, complete with two diametrically opposed rational strategists
plotting the endgame of the other.36 It is no surprise that the Soviet Union,

9800.indb 175 6/2/16 3:05 PM

176  Chapter 5

which reigned as chess hegemony for most of its existence, took its strategic
chess, computer, and long-term planning thinking seriously. Among those
thinkers stands Mikhail Moiseevich Botvinnik, who, although not quite
the brightest star in the constellations of Soviet grandmasters, is nonethe-
less remembered as the patriarch of Soviet chess for innovating and insti-
tutionalizing rigorous systems for gameplay. As this section explores, with
the support of Glushkov and others, Botvinnik even programmed his own
end game for cold war chess itself. His Pioneer Project stood as an attempt
at computer chess programming that he felt would bring the Soviet Union
one step closer to triumph in strategic political economic planning.
Raised in St. Petersburg, the son of a dental mechanic who had earned
the right to move beyond the Pale of Settlement, and married to a ballerina
(the other superior Soviet art of elegant maneuvers), Botvinnik (1911–1995)
came to chess at the late age of thirteen and left the chess world a differ-
ent place seventy years later.37 In 1935, at age twenty-four, he became the
first Soviet grandmaster, and by 1957, under his guidance, there were nine-
teen Soviet grandmasters, with roughly twenty new masters emerging every
year. As a figure astride Soviet chess history, Botvinnik is remembered today
for establishing “the Soviet school of chess”; for mentoring world-famous
chess figures Anatoly Karpov, Vladimir Kramnik, and Gary Kasparov; and
for promoting disciplined chess training (a cross of physical and mental
exercise). He was also a theorist of long-term strategic planning who pio-
neered Soviet computer chess and chess schools based on his notational
system for recording chess play. That notational system preceded what is
now known in the chess world as “the Book.” A fascinating character on
his own right, Botvinnik figures here because he is an early Soviet network
visionary. Like the other distinguished scientists and long-term strategists
who were committed to the Soviet way of life, he proposed that the state
use computers to optimize and resolve its long-term planning problems in
economic and political spheres.
Botvinnik’s combination of professional success and political notability
was a rare distinction for advanced Soviet chess players, whose demanding
careers as civilian celebrities rarely left time for anything else. He even was
awarded a national medal of honor for his work as an engineer at the same
time that he was establishing himself as a world chess grandmaster. In 1954,
six years after defeating the reigning American to win the world champi-
onship, Botvinnik came as close to a public icon as the Soviet Union had
then (the superstardom of cosmonaut Yuri Gagarin came later). Botvinnik
received spontaneous standing ovations on entering movie theaters and
was one among few other than members the Party elite who had a private

9800.indb 176 6/2/16 3:05 PM

The Undoing of the OGAS, 1970 to 1989  177

car and driver. Buoyed by such a reputation, he wrote a strong-willed let-

ter to Pravda in 1954, the year after Stalin’s death, detailing a long-term
strategy for world domination without having to go to nuclear war. His
suggestions involved calculated moves and countermoves through which
the socialist leaders would grant the masses of petty capitalist owners their
material wealth in exchange for their acceptance of the socialist revolution
without atomic combat.38
For such a brazen public stunt, the political secretariat rebuffed him and
threatened to throw him out of the Communist Party. In the 1960s, he
publicly repented and avowed his Communist credentials by publishing
Computers, Chess, and Long-Range Planning, which describes how the domi-
nation of the Soviet school of chess over the Americans was an expression
of superior long-range socialist planning. In 1968, having been influenced
by Claude Shannon’s less well-known 1950 work on computer chess, Bot-
vinnik published An Algorithm for Chess, which successfully demonstrated
how to algorithmically organize attacks against an opponent’s position
from challenging tactical positions. Even though his algorithm excelled in
solving technically stressful positions, it also had the frustrating tendency
to overlook the simplest tactical moves.39
Backed by major cyberneticists and computer engineers including Vik-
tor Glushkov and his colleagues—such as Bashir Rameev (who developed
the Ural computer series), Viacheslav Myasninkov, and Nikolai Krinitskiy
in the 1970s and 1980s—Botvinnik poured his energies into what he called
(drawing on the Stalinist vocabulary of his youth) the Pioneer Project, a
computer chess program that was designed to imitate how the brain of a
grandmaster works.40 The OGAS Project algorithms were designed to ignore
the bulk of all computationally possible moves and instead to concentrate
on the most probable moves. Attempts were made to develop an algorithm
with a long-term intuitive “feel” for the board. The brute force approaches
(which calculate all possible moves in branching decision trees) eventually
won out with the arrival of faster computers in the 1970s, outpacing Botvin-
nik’s selective but theoretically more sophisticated approach. Nonetheless,
his prodigies in training—Kasparov among them—remember their surprise
at hearing that Botvinnik was confident that his selective program would
one day consistently beat them all. Nevertheless, it—much like Glushkov’s
attempt to develop intuitive macroprocessing and natural language pro-
gramming that mimicked the neural processing and speech patterns—bore
fruit in other spheres of application. For example, Botvinnik, in his career
as an electrical engineer, reconfigured his Pioneer algorithm into planning
maintenance repair schedules for power stations across the Soviet Union.41

9800.indb 177 6/2/16 3:05 PM

178  Chapter 5

The Pioneer Project and the OGAS Project shared more than a common
organizational framework and set of state-of-the-art computers. At core,
they shared a commitment to organize the real-time management of scarce
computational and economic resources. In the early 1980s, Botvinnik tried
to salvage the national economy with another proposal that he sent to
Party leaders. It contended that the Soviet economy should be regulated
by a software program that, like his Pioneer chess algorithm, would take a
generalizable approach to reasoned decision making. Botvinnik thus stands
out as the last of major Soviet figures (with Kitov, Kovalev, Fedorenko, and
Glushkov) to propose using computer software to salvage the command
economy. Available records do not speak to his proposal’s reception, except
that it was rejected at the highest levels. By the mid-1980s, Gorbachev’s
reforms had already sufficiently introduced market elements into the for-
mal command economy to render impossible any future systematic man-
agement of the economy. In the early 1990s, shaken by the collapse of the
Soviet Union and years from death, Botvinnik reached out one last time
with strategic advice for Yeltsin’s government but to no avail.42
There is a truism in the history of science that science serves many spe-
cific social purposes but basic research need not begin with any single goal
in mind. Biologists, for example, run test on fruit flies—or Drosophila—not
because they are particularly devoted to improving the life of fruit flies;
they do so because fruit flies are convenient test subjects that reproduce
quickly and cheaply. Computer chess has been called “the drosophila of
artificial intelligence” (Alexander Kronrod’s phrase, popularized by Ameri-
can computer scientist John McCarthy) because it is thought to stand in
as an affordable test case for larger strategic programming projects, which
include both artificial intelligence as well as planning the Soviet com-
mand economy.43 Kronrod, himself a distinguished Soviet mathematician
and computer scientist, also collaborated with Kantorovich on the com-
puter planning of the economy and with Botvinnik on the algorithm that
defeated the Kotok-McCarthy American chess program in 1966 and 1967.44
The unexpected joy of computer programming lay in finding new applica-
tions for old techniques, which in many ways was the same allure that
fascinated general-purpose computer programmers since Turing. Although
OGAS, EGSVTs, ESS, ESAU, and even Botvinnik’s Pioneer Project “failed” on
their own terms, they also should be remembered for their contributions to
ongoing macrolevel experiments in rational planning, administration, and
policy making in a world of global information networks.
As these cases suggest, the consequences of the current networked infor-
mation revolution cannot be easily anticipated. The chess community has

9800.indb 178 6/2/16 3:05 PM

The Undoing of the OGAS, 1970 to 1989  179

long concerned itself with the advancement of computer chess programs,

which grew exponentially more sophisticated from 1970 through 1989,
when Gary Fields was first defeated, and then again until Deep Blue’s con-
troversial victory over Kasparov in 1996. Since at least Wiener in 1964, crit-
ics have contended that superior computer chess programs were inevitable
but would diminish the value of chess as a human activity.45 The situation
led early computer chess critics to bemoan their state with the defeatism of
the final scene in the 1980s film War Games. After all, if the heirs of Botvin-
nik’s Pioneer program dominate the best players today, the human will to
be the best has already been undermined. What is the point of playing, the
chess enthusiasts worried, when everyone loses every game?
Such handwringing by chess purists against the artificial intelligence
community has since been sidelined—and by neither the triumph of tech-
nology over humanity nor the triumph of humanity over technology. Chess
as a human pastime has not dwindled in the face of virtually indomitable
computer programs. Instead, networked computers have sped the spread
and growth of the global chess community. The number of online human-
to-human and human-to-computer chess has exploded since Kasparov’s
defeat for unrelated and seemingly mundane reasons. No longer encum-
bered with the burden of serving as a shadow stage for cold war intrigue,
long-distance chess in real time over computer networks is now an every-
day reality.
Botvinnik’s influence on networked computers and chess continues to
surprise. It is not Botvinnik’s sophisticated computer algorithm but his
foundationally basic notational system that has had the most lasting effect
on the now globally networked game of chess. Thanks to well-codified chess
notation systems that were popularized by Botvinnik, computer record-
keeping capacities have allowed millions of games of top-level chess to be
catalogued into a database known as “the Book.” Recently, for example,
a German company named ChessBase has been scrutinized for its widely
used database of chess moves that organizes prior games, new move oppor-
tunities, and errors in human play, which effectively reduces chess games
to enormous decision trees of known and unknown pathways of game
progression. Critics have accused its founder, Frederic Friedel, of having
“ruined chess” because few games now occur that include new combina-
tions of moves that are not found in “the Book.”46 The result is a new cold
war tension of human players against the book, in which top chess players
and their opponents know that, given almost any chess board arrangement,
the best game they can play is played out in “the Book.” Botvinnik’s secret

9800.indb 179 6/2/16 3:05 PM

180  Chapter 5

library of index cards that recorded global grandmaster games, saved exclu-
sively for study by students at his Soviet school, did not migrate online, but
it modeled what has become the global networked norm. Top players world-
wide now memorize tens of thousands of recorded games and positions.
All chess competitors now play aware of the networked heir of Botvinnik’s
book and the humbling fact that most chess sequences have already been
played before. The introduction of networked computing is driving a curi-
ous situation (however common when new media become mainstream) in
which Botvinnik’s dream has now been achieved (for example, since 2005,
the best software programs routinely trounce the best humans at chess)
without appearing the affront to humanity its critics predicted it would be.
In fact, global communication networks have made correspondence chess
(with humans and computers alike) more popular. Perhaps the enduring
attraction to strategic pastimes reveals, with a gesture to Walter Ong, that
there may be nothing more human than artifice. (Consider the complex
rules and recipes behind baseball and apple pie.)47
Computer cold war chess offers a view of the historical preoccupation
with global and long-term planning strategies from Liebniz to modern-
day generals.48 The reformist efforts of Kantorovich, Glushkov, Fedorenko,
Kovalev, Kharkevich, Botvinnik, and many others are not exceptional.
Rather, the introduction of the digital network in socialist cybernetic plan-
ners and the sharing of “the Book” in chess underscored something that was
at once peculiar yet normal. Networks make knowledge generalizable or at
least generally shareable and remixable—whether a dataset shared by net-
work or a playbook shared within a chess school. The consequence of that
record, after it was repurposed from the secret index files of Soviet libraries
into open-access public repositories, in turn is purported to do nothing
less than remake the chess world. Such a grandiose sentiment outlines the
strong intellectual affinity between Soviet cybernetwork visionaries and the
modern preoccupations with network-enabled public recordkeeping and its
automated extension, surveillance.
Simultaneously, the experience of Soviet computer chess also under-
scores the critical fact that, although military and civilian projects in the
Soviet Union suffered from being strongly separated, the cold war culture—
especially cybernetic tools, game theoretic strategic thinking, and the com-
putational management of limited resources—has spread the influence of
military and strategic thinking far and wide into everyday matters of poli-
tics and economics. In chess as in planning, the separation of military and
civilian administration offers no guarantee of the same in modern society.

9800.indb 180 6/2/16 3:05 PM

The Undoing of the OGAS, 1970 to 1989  181

How Hidden Networks Unravel Cybernetworks

This chapter has introduced and advanced an argument based on the infor-
mal character of the Soviet system outside of the centralized military com-
mand. The dynamic vitality of the system—unregulated competition with
unpredictable promotions and demotions—did not always benefit the well-
positioned and talented network entrepreneurs and system reformers, as a
number of case studies have shown. As Kitov’s Red Book show trial dem-
onstrates, military superiors were free to punish their own best and bright-
est for suggesting that networked computing capacities should be shared
beyond narrow military applications. His superiors formally accused him
not of displaying generosity toward civilian concerns but rather of going
outside formal military communication channels, which underscores the
depth of the structural military-civilian divide behind the Soviet network-
ing story. The early partnership between Glushkov’s Institute of Cybernetics
and Fedorenko’s Central Economic-Mathematical Institute illustrates some-
thing of their double-edged situation. Glushkov and Fedorenko faced oppo-
sition from the centralized military command of the Ministry of Defense,
which denied them access to military networks and the institutional knowl-
edge base that supported those military networks. At the same time, they
also faced a more subtle institutional obstacle to OGAS that came from the
civilian economic sector. The unpredictable currents and institutional drift
of the state bureaucracy, including a flush of untethered funding, pulled
their young, growing, and capable research staffs in divergent directions—
including a focus on macroeconomic reform under Glushkov in Kiev and a
focus on microeconomic reform under Fedorenko in Moscow.
Near the end, Glushkov reflected on the sources of the obstacles that
his team faced when they were developing the OGAS, the EGSVTs, and
national economic reform. His sense of disappointment with his own gen-
eration was particularly acute, and his last book that was published while he
was still alive targeted schoolchildren as its audience—What Is the OGAS?49
In 1983, while on his death bed and suffering from an apparent tumor
of the medulla, Glushkov proclaimed that the OGAS was his “greatest life
work,” after over twenty years of dedicated effort and a long list of signifi-
cant accomplishments in other major scientific fields.
During the last nine days of his life, while constrained to his hospital
bed in the Kremlin, Glushkov insisted on working—just as he had done
back in his hospital bed in 1962. During those final days he dictated his
life memories to his daughter Olga and received as a guest the deputy of

9800.indb 181 6/2/16 3:05 PM

182  Chapter 5

Dmitry Ustinov, one of his staunchest supporters in the military. Ustinov,

Glushkov reflected, had managed to do in his military career what Glush-
kov could not do in the civilian sector—rise from the chair of the Supreme
Council of the National Economy under Khrushchev to wield power to
reform the Ministry of Defense as its minister and marshal of the Soviet
Union. Ustinov’s deputy listened to the dying man’s account of the “long
ordeal” of his constant skirmishes with state bureaucracy before asking
what the minister of defense could do to help. Glushkov, wrapped in the
tubes of respiratory support, sat up and growled a memorable deathbed
witness to military might and its remove from civilian concerns, “Let him
send a tank!” Before an excessive growth in his own nervous system could
bring down this champion theorist of the Soviet economic nervous system,
Glushkov tried to comfort his grieving wife, Valentina. In his hospital bed
in the Kremlin, he turned to her and spoke about the possibilities of immor-
tality: “Be at ease,” he said. “One day the light from our Earth will pass by
constellations, and on each constellation we will appear young again. Thus
we will be together forever in the eternities!”50
After Glushkov died on January 28, 1982, the OGAS vision continued
to radiate outward and did not immediately fade from state discussions,
social networks, and print media. Anatoly Kitov attempted to reanimate
the proposal by writing directly to General Secretary Gorbachev in October
1985. Kitov, then the chair of the Department of Information Technology
at the Plekhanov Moscow Institute for the National Economy (part of the
Russian Academy of Sciences), recounted the history of the OGAS Project—
the scattered development of unconnected ASUs in the 1960s and 1970s,
Kitov’s repeated appeals to the state for support, the subsequent disappoint-
ment with the spread of ASUs and the potential for networking them, the
lack of state coordination over technical as well as administrative matters
(especially the cooperation problem among separate ministries), and the
fact that “we do not have modern reliable personal computers.” “I think
that this report constitutes an objective analysis of the last thirty years of
developing information technology,” Kitov concluded in the letter: “may it
bring specific benefit and capacity for further decisive action.”51
This time, Kitov’s 1985 letter was not intercepted, although his reclama-
tion of the OGAS situation came to the same effect as his Red Book letter
did almost thirty years earlier: nothing would be done. The way that he
was told this, however, reveals a crucial look into the inner workings of
the administrative state. On November 11, 1985, Kitov received a phone
call from Yu. N. Samokhin, a representative from the economic division
of the Central Committee that had reviewed his letter to General Secretary

9800.indb 182 6/2/16 3:05 PM

The Undoing of the OGAS, 1970 to 1989  183

Gorbachev. Kitov’s notes record that he was told two things: first, he was
to be thanked for his contributions, and second, “not everything in the let-
ter is supported by the economic division.” The Politburo and the Central
Committee, he was told, “had other functions, not those of the automatic
management of the command economy.” The Politburo was already sup-
porting the creation of a state committee of information technology, and
at the moment, that, not the economy, was the state’s priority. Kitov, at the
end of the telephone conversation, asked to receive the reply in writing and
was told that the Central Committee did not provide written replies.52 The
likely reason for not offering a reply in writing was that the Central Com-
mittee did not want to proliferate in writing its own contradictions—in
this case, that the economic division of the governing body of the Soviet
state does not concern itself with the automatic management of the econ-
omy. No doubt Kitov felt that this reply was begging the question: that, it
seemed, had been precisely the problem all along.
Such telephone revelations, however, did not keep the state, one year
later in 1986, from pronouncing with the force of law that the economy
actually would pursue the following Glushkovian demands over the com-
ing five years (in the twelfth five-year plan): it would double the level of
automation, organize the mass production of personal computers, increase
the installation of computers by 100 to 130 percent, build computer centers
for collective use, create integrated information banks, and significantly
increase research in information theory, cybernetics, microelectronics, and
radio physics.
The passage of time has allowed some reflection on the sources of
these challenges. In 1999, Fedorenko contemplated the stubborn fact that
decades of CEMI efforts to develop macrolevel economic models had born
very little fruit in part because “the problem was too multidimensional and
multifactorial.” But “the very hardest,” Fedorenko admitted without clari-
fication, “was the ‘human factor.’”53 Three decades earlier, the problem was
effectively the same. In 1968, Kitov summarized his own frustrations in a
personal letter to Lyapunov, not so much as a problem of human personali-
ties or specific personnel but as a problem of cultural resistance to reform
in the institutions:

The top leadership realizes the importance of [the introduction of computers into
the national economy] but takes no effective measures in support of such work,
while responsible officials from the ministries and other government agencies … dis-
play no interest in the automation of management for the optimization of planning.
The problem is apparently rooted not in their personalities, but in their positions
and in the overall traditions, which change very slowly.54

9800.indb 183 6/2/16 3:05 PM

184  Chapter 5

Here Kitov faults not the top leadership of the country but the insti-
tutional logic of administrators in the middle levels of ministries. Kitov’s
complaint holds that middle managers habitually reaffirmed the status quo
(as is also the case in many large organizations) and that the dynamics
within such conservative institutions were paradoxically nonsystematic. As
countless Soviet officials have observed: “Having different ministries is like
having different governments,” and the battle between civilian ministries
often flares into nothing less than internecine civil war.55
The paralyzing competition between these dynamic, unregulated min-
istries constrained the possibilities of both systematic institutional growth
and purposeful reform. Interministry cost-sharing and cooperation rarely
happened. Whenever high-ranking Soviet administrators wanted to pro-
mote a major project (such as a national network), the primary avenue
for action available to them, as Garbuzov’s counterproposal anecdotally
indicates, was to create an entirely new institute within preexisting admin-
istrative silos. Thus, the attempt to create a supervisory institute for a par-
ticular sphere of responsibility (such as finance, statistics, or the OGAS)
created intractable points of competition between those institutes. Instead
of easing the conflict among administrative standards, every new umbrella
institution introduced a new competitor and exacerbated the power skir-
mishes. The attempt to create a hierarchical bureaucracy to resolve conflicts
of administrative interest often generated more, not fewer, opportunities
for infighting among neighboring bureaucracies. So the chasm between
military and civilian administrations was perhaps not entirely insurmount-
able: while the military kept the country ready for war with the enemy,
the civilian bureaucracy was already at war with itself. Unable to receive
the same preferential state treatment as the military, the Soviet economic
bureaucracy militarized itself against itself.
The uneven economy of those administrative silos often pivoted around
surprisingly few well-placed administrators and veto points. Consider, for
example, that the tenures of the chair of the Council on Cybernetics Aksel’
Berg and the mathematician Mstislav Keldysh, president of the Academy
of Sciences of the Soviet Union (1961–1975), coincided with the rapid
growth of Soviet cybernetic academic preoccupation in the 1950s and 1960s.
Together, Keldysh and Berg personally facilitated the creation of all four
main institutes featured previously, including the Computation Center 1
that Kitov directed following his optimistic report about the future of com-
puting technology in 1953, the Institute of Cybernetics in Kiev under the
directorship of Viktor Glushkov (1962), the Central Economic-Mathematical
Institute in Moscow under the directorship of Nikolai Fedorenko (1963), and

9800.indb 184 6/2/16 3:05 PM

The Undoing of the OGAS, 1970 to 1989  185

the Institute for Telecommunications in Moscow under the directorship of

Aleksandr Kharkevich (1963).56 Given the immense reach and corresponding
tangle that is Soviet cybernetics, a disproportionate amount of its adminis-
trative growth took place at one or two degrees removed from these men’s
signatures and oversight. In the institutional growth period of Soviet cyber-
netics from 1953 to 1964, the roles played by supporting administrators like
Keldysh and Berg helped extend the argument that Soviet institutions often
experienced periodic spikes of exceptional growth followed by long periods
of underdevelopment.57 Perhaps the most striking record of the explosive
state imagination for computing technology, at least as of 1963, that pro-
pelled Soviet cybernetics, CEMI, the Institute of Cybernetics, and the associ-
ated OGAS Project into the mainstream of Soviet political system is in the
recently uncovered Party resolution published on May 21, 1963. This resolu-
tion, issued by both the Central Committee and the Council of Ministries,
declared that the Soviet state would advance nearly twenty nationwide new
or transformed tasks and institutions involving computing technologies,
including the reform by computer network of the command economy.
Even so, the economic bureaucracy proved less resolute about embracing
such sweeping technological reforms. As a house divided, the bureaucracy
was unpopular with practically everyone, including the underserved pub-
lic, scientists like Glushkov and Liberman, and politicians (such as Mika-
hil Gorbachev) who publicly ran against the bureaucracy not in earnest
hope of reforming it but to ensure their own political popularity with the
public.58 In the main, the old guard of orthodox planners who adminis-
tered the system benefited from it, although it would be a stretch to say
that they approved of how it functioned. Crisscrossing structures, personal
favors, and impartial administrative reforms plagued the hierarchy that
held together the national, regional, and local planning committees. This
ensured that, despite the state’s approval of a single national plan, there
were as many contested plans as there were administrators of the single
plan. (No plan can plan away its own private interests.)
By the time that Gorbachev’s perestroika and glasnost policies were
introduced between 1985 and 1989, the national economy could no lon-
ger mobilize around mathematical economic reforms. Official statistics
hold that between 1986 and 1988 the economy grew by 2.8 percent and
in 1989 by 2.4 percent, although in practice real economic progress, like
official economic statistics, was not meaningful. By 1985, after perestroika’s
decentralizing reforms (according to M. S. Shkabardni, one of Glushkov’s
colleagues), the idea of an economy that was rationally decentralized by
OGAS “interested no one. Everyone had forgotten about it. No one even

9800.indb 185 6/2/16 3:05 PM

186  Chapter 5

thought about it.”59 The Party leadership had many more pressing worries
to consider, including the high capital investment that would be required
if the economy were to be reformed by networked computers amid a ris-
ing stream of affordable personal computers from the West. The OGAS had
long appeared a prohibitively expensive “hero project,” but now even the
more modest EGSVTs network could be built out by individual citizens who
were working on Western computers. In a sense, that is what happened:
large state network projects were abandoned, and in the late 1980s, a few
Soviet citizens joined in purchasing and connecting personal computers
to globalizing communication networks. By 1989, private Soviet citizens
began logging onto early Internet chatrooms, by which time the OGAS
Project, like the Soviet state, was slipping into history.
In summary, the OGAS Project was shipwrecked on the capricious
unregulated conflicts of self-interest that occupied the civilian knowl-
edge base (including but not limited to the economic bureaucracy) of the
Soviet system. It fell prey to the conflicts of interest that it sought to set
aside with automated networks. The sources of those conflicts arose from
the yawning disconnect between the formal plan for the civilian sector,
which was clearly hierarchical, and the massive gray economy of informal
exchange and personal favors. Each layer of the command economy—the
national, regional, and factory planners and managers—benefited from a
slack and informal freedom that allowed them to solve problems outside
of the plan’s commands. By rationalizing, making explicit, and automating
those resources, Glushkov’s vision directly opposed the informal economy
of mutual favors that oiled the corroded gears of Soviet production. In the
end, the OGAS Project fell short because, by committing to rationalize and
reform the heterarchical mess that was the command economy in practice,
it promised to encourage the rational resolution of informal conflicts of
interest—which worked against the instinct to preserve the personal power
of almost every actor that it sought to network.


The portrait of the final chapter of the OGAS Project that is presented here
fills out and begins to complicate the conceit with which this book began—
that global computer networks arose from collaborative capitalists, not
competing socialists (or in light of the OGAS Project, not from the unregu-
lated conflicts of self-interested socialist institutions). This surely is no plain
victory for any political order, nor is it only a plea for virtuously regulated
market-state interactions. Self-interest has been a recognized engine of

9800.indb 186 6/2/16 3:05 PM

The Undoing of the OGAS, 1970 to 1989  187

human behavior since at least the ancient Greeks and found in any eco-
nomic order. (I understand self-interest here to be an ambiguous quality
that is more basic than any particular economic order. It can range from a
virtue as distinct from selfishness as satisfaction is distinct from hedonism
to, as the Gautama Buddha taught, a signal vice of enduring dissatisfac-
tion in life.)60 The Soviet socialism that the Project sought to reform never
worked as it planned in part because of the economic administration’s
mismanagement of its own conflicting internal egotisms and mutinous
ministers. Its political economic tragedy lies in the flooding of the gray
economy with the informal self-interests that the planned interests of the
command economy—especially a technologically rationalized one—could
never accommodate. It was not the absence but the presence of vibrant
unregulated markets of conflicting forces driven by self-interested admin-
istrators that kept the Soviets from networking their nation and command
economy. In another sense, the Soviet networked command economy
fell apart not because it resisted the superior practices of competitive free
markets but because it was consumed by the unregulated conflicts among
institutional and individual self-interests—including the institutional rival-
ries that sprung up between Glushkov and Fedorenko’s competing efforts
to network and model the economy, the ministry mutiny over funding
between the Central Statistical Administration and the Ministry of Finance
over the network plans, and the adhocracy of the Politburo.
There is a problem, however—not with the history but with the ends
of such critical analysis that inverses the role of regulated capitalist states
and unregulated socialist economies. In so doing, it recapitulates the liberal
economic coordinates for imagining the state as the site for public inter-
ests and the market as the site for private interests. The conclusion to this
book outlines several reasons that such an analysis, although tempting,
cannot hold on its own. Before that conclusion, let us summarize a few
larger points that previous chapters have built toward.
First, the Soviet economic system did not work—except for when it
did, which was mostly for highly centralized militarized projects. It is rea-
sonable to presume, as social scientists and cyberneticists alike have been
doing, that the Soviet formation of socialism cannot be separated from the
economic and political woes that arose due to underlying structural con-
tradictions. For the most part, those contradictions have been framed in
terms of private (usually market) interests that were in competition with
public (usually state) interests. Given this framework, the history outlined
above may prompt defenders of private (market) interests to offer remind-
ers about how, in the Soviet Union, private and public sectors managed at

9800.indb 187 6/2/16 3:05 PM

188  Chapter 5

best an “uneasy coexistence” or about how four decades after collectiviza-

tion, private (market) plots that comprised 3 percent of Soviet agricultural
lands managed to produce nearly 30 percent of the gross value of Soviet
agriculture.61 To argue that market solutions would work better does not
begin to describe or distinguish what I believe the OGAS Project history
reveals to be the depth and range of private interests at work in human exis-
tence. It is just as easy, in what prominent economist Igor Birman endorsed
as Soviet “anecdotal economics,” to list examples of how the same kind of
private self-interest that put bread on the table of starving peasants also cor-
rupted socioeconomic life elsewhere. Anecdotes from everyday economic
life relate that 80 to 85 percent of gasoline, according to some estimates,
turned up on the black market;62 construction workers built new apartment
buildings to state specifications but refused to connect the toilets to the
sewage systems until vzyatki and podkupki (bribes) were paid; maternity
nurses extorted 200 ruble notes from birthing mothers before using a sterile
needle and anesthetic; grieving families had to pay 2,000 rubles to bury
their mother, despite the guaranteed “free” state funeral and burial. The
fact that most numbers were anecdotal suggest how actively corrupt Soviet
economic life already was.
Self-interested corruption is so much a feature, not a bug, of Soviet eco-
nomic life that it cannot be the result of market absence or state failure
alone. Anecdotes of administrative cunning (not incompetence) abound.
In Grossman’s phrase, “the Four B’s: barter, black market, blat, and bribe”
summarize the economic engine of Soviet self-interest run amok.63 An
entire biscuit factory once went underground in Georgia, producing four
times its planned quota through hidden informants, bribery, and social
screens;64 a seat on the trade committee in Moscow sold for 50,000 rubles
in 1990 (and current prices for other positions can be found online today);
and Central Committee members filled foreign bank accounts by extracting
bribes from officials in the trade ministries.65 The Soviet joke puts it well:
Brezhnev is showing his mother how well he’s done, and he shows her his
suite in the Kremlin, his dacha in the country, his Black Sea villa, and his
Zil limousine. “All very nice, dear,” she says. “But what will you do if the
Bolsheviks come back?”66
These anecdotes constitute what we might call revolts in miniature. They
are an expression of private unrest—of local resistance to a society whose
public institutions did not have to serve the public. A liberal economic
analysis to these problems might describe the informal networks of com-
peting private interest as variously productive or rent-seeking, depending
on whether the activity at hand created or depleted economic resources.

9800.indb 188 6/2/16 3:05 PM

The Undoing of the OGAS, 1970 to 1989  189

Varied critics of the Soviet economy have interpreted the collapse of the
public interests of the state and the private interests of the market into the
command economy to be a hallowing out of means for Soviet citizens to
seek their own self-interest through formal mechanisms.67 Consequently,
informal means, whether creating islands of penny capitalism or engaging
in systematic corruption, are all that is available.68 The liberal economic
critique accuses the public state of systematically smothering and driving
underground private self-interest. Applied to the OGAS case, the standard
critique follows: the OGAS Project could not hope to reform the command
economy because its very purpose ran counter to the interests of those who
held hostage that economy in need of reform.
The argument advanced in the conclusion to this book seeks to go one
step further. It seeks to rearrange our thinking about cold war networked
culture by twisting the standard liberal economic distinction between
public states and private markets to feature a classical distinction between
public polis (community) and private oikos (household). Instead of seeking
to place blame on either the state for publicly stifling private self-interest
or the individual bureaucrats for seeking to protect their professional self-
interest by opposing reform projects, I suggest that the OGAS history reveals
a third approach to social reform. The OGAS Project sought technocratic
reform that is both public in its relationship to the market and private in
relationship (or privy) to the state. It does not matter whether one faults the
public state or the private market elements in the command economy that
the OGAS Project tried to reform because they belong to the same classical
category of private interest. Both state and market actors, collapsed into
the Soviet command economy, sought their own private self-interests with
certain consequences for how social and technological networks shape one
As the Soviet network stories show us, cold war economic orders prove
more compatible in practice than in liberal economic theory, if for noth-
ing other than their shared liability to collapse without careful regulation.
Neither American-style capitalism nor Soviet-style socialism should be
considered a sufficient philosophical banner for making our way into a
networked world. If there is a shared baseline, it must be found in the agree-
ment to regulate and restrain self-interest that is common to the visions
of both Smith and Marx. The social necessity of restraining self-interested
competition unites, not divides, the modern legacy of cold war socialism
and capitalism. The following conclusion explores a few consequences for
reintroducing a search for the role of public interests, in a classical sense of
the term, in Soviet as well as contemporary network worlds.

9800.indb 189 6/2/16 3:05 PM

9800.indb 190 6/2/16 3:05 PM


This is the story—told for the first time in any language in book form—of a
particular path not taken into the modern network age. Soviet scientists—
led by Viktor Glushkov and his OGAS team between 1959 and 1989—could
have developed a computer network project that brought about significant
political, economic, and social changes. Had they done so, the current
global network culture could have looked very different. Why did these
network entrepreneurs not succeed? On what factors did the tragic twists of
the tale we might dub the Soviet “InterNyet” hang?1
Faced with a struggling command economy, attempts to revitalize Soviet
cybernetics, and a search for societal reforms after Stalin’s bloody gover-
nance, Soviet researchers proposed as early as 1956 that computers should
be used to control economic decision making. No one proposed that these
computers be connected, however, until Anatoly Kitov, the military scien-
tist who had “discovered” cybernetics in 1952, proposed in 1959 that civil-
ian economists use existing military networks to solve economic problems,
for which suggestion he was promptly dismissed from the army. At the same
time as Kitov was making short-lived network proposals, Gluskov teamed
with him and others to propose in 1962 a complex three-tiered hierarchi-
cal computer network that would transfer economic information along as
many as, in its most ambitious proposal, twenty thousand local computer
centers, several hundred regional centers, and one central computer center
in Moscow. Over the years, this prohibitively expensive proposal was scaled
down (and back up) to match the political climate. Nevertheless, the goal
of this interactive, remote-access network remained the same—to reduce
the coordination problems that had long beset the command economy.
On and off over the next twenty years, Glushkov’s OGAS team met resis-
tance from at least five groups: (1) the military wanted nothing to do with

9800.indb 191 6/2/16 3:05 PM

192 Conclusion

civilian affairs, especially when that meant fixing the command economy
that already fed its coffers; (2) the economic ministries (particularly the
Central Statistical Administration and the Ministry of Finance) wanted the
OGAS Project under their control and fought to the point of mutiny to keep
competing ministries from controlling it; (3) the bureaucrats administering
the plan feared that the network would put them out of a job; (4) factory
managers and factory workers worried that the network would pull them
out of the informal gray economy; and (5) liberal economists fretted that
the network would prevent the market reforms that they sought to intro-
duce. Instead of a national network, dozens and then hundreds of local
computer centers—or automated management systems (ASUs) were built
in the late 1960s and 1970s, although they were never connected. Thus the
dream of networking Soviet socialism into a brighter communist future did
not come to pass. This conclusion remarks on why this never happened
and then hazards a few concluding comments and pronouncements.
There are many reasons why there were no such Soviet networks. But
first is a reason to care about this story. Soviet network history invites us
to think about the historical conditions of national computer networks
without the assumptions behind the rise of current global digital networks.
In other words, the OGAS story is a test case in how network projects
could have developed in societies that were not preoccupied with markets,
democracies, and personal liberties. Network projects without political and
economic liberal values are not condemned from the start. Instead, after
these cases are examined on their own terms, they can help control for,
challenge, and rethink the conditions of possibility that are assumed to
govern digital global networks. The Soviet network projects did not fail
because they did not possess the engines of particular Western political or
technological values. They broke down for their own reasons.
And these reasons were not the popular Western misconceptions. The
standard criticism of Soviet technological backwardness (technological
“behindness” would be more accurate) cannot describe on its own what
prevented Soviet civilian networks from developing because the Soviet mili-
tary possessed functioning long-distance computer networks since the mid-
1950s and local area networks were linking ASUs since the mid-1960s. The
technical know-how was in place. Nor can it be that computer networks are
somehow inimical to closed cultures because computer networks have been
serving military, authoritarian, and cybersecurity cultures for decades. That
said, the history of Soviet technology overflows with technical problems—
such as a lack of interoperable hardware or software for ASUs. Almost never,
however, does the root explanation for Soviet technological problems lie in

9800.indb 192 6/2/16 3:05 PM

Conclusion 193

sheer technical incompetence. This advanced superpower state provided

strong support for science.2 The root problems with technology are any-
thing but technological.

Beyond the Binary: Arendt and OGAS

Why was there no Soviet Internet? This book holds that leading Soviet
scientists and their supporters—especially the OGAS team lead by Viktor
Glushkov—tried repeatedly but could not network their nation with com-
puters due to entrenched bureaucratic corruption and conflicts of interest
at the heart of the system they sought to reform. McCulloch gives us a
fresh term: heterarchies of conflicting private interests stalemated virtuous
attempts to reform the hierarchical economic bureaucracy. If the Internet is
not a thing but an agreement, as the phrase goes, perhaps the Soviet Inter-
net is not a thing but a disagreement. (There is often more to learn from the
latter than the former.)
This thesis, which expands on the standard interpretation, can be taken
further. The history of the OGAS Project is akin to the history of a miscar-
ried effort to perform an IT upgrade for the corrupt corporation that was
the USSR itself. USSR, Inc., in other words, functioned as the world’s larg-
est corporation, and its private interests were internal market capture, the
avoidance of the transaction costs of the capitalist market, and the concen-
tration of power to itself. The political need for the OGAS Project appears to
represent the grander inability of the hierarchical state structure of socialist
politics since Marx to build and sustain innovation and reform in the age of
industrial and information capitalism that the Soviet Union straddled. The
network reform effort did not take into account its own effects on the for-
mal command economy because the OGAS Project ran against the private
interests of those who governed within an informal mixed economy. The
perpetual conflict of self-interests that were internal to the Soviet system
helps describe the continuous institutional tumult, frequent and ineffec-
tual reforms, and currency of informal influence that underwrote the sup-
posedly staid Soviet bureaucracy. The root problem here appears to be not
the cold war binary between international economic systems but the binary
that was internal to the Soviet economic system. Hidden, informal, and
often vicious administrative networks prevented public, formal, and poten-
tially virtuous computer networks from taking the Soviet Union online.
This view that the Soviet Union can be understood as a corrupt corpora-
tion also has its limits. In theory, it reads Soviet network history as it would
read a Western state. In practice, it risks using the liberal economic values of

9800.indb 193 6/2/16 3:05 PM

194 Conclusion

market, state regulation, and individual interests to criticize socialist values

of state-managed economies and collectivized interests—in effect, rehears-
ing the very political economic divide that it seeks to revise. This view falls
short of explaining the motives and behaviors of other relevant actors.
Although it depicts the perspectives of both the internal reformer (espe-
cially the scientists and administrative supporters of the OGAS Project) and
the external critic, the interpretation does not describe why the militarized
state, economic bureaucracy, and citizen workforce actively opposed ideo-
logically faithful network projects. Why were their private interests in play
at all, how can that question be described without rehearsing the exhausted
cold war showdown between markets and states, and how might our answer
to that question help focus critical attention on the contemporary scene?
Let us tweak our terms to state the situation more clearly. The OGAS Proj-
ect could not achieve its end goal of reforming the Soviet economy because
the hulking households of private power—the military, the corporation,
and the state—compelled it into serving their private economic, not public
political, interests. Consider the language of Hannah Arendt’s The Human
Condition—a landmark work of political theory that introduces its disen-
chantment with normative liberal values with a discussion of Sputnik and
the nuclear age, the two ingredients that, once combined, could spell instan-
taneous planetary annihilation. For Arendt, the distinction between the pub-
lic and the private is not the liberal economic opposition of the public state
and the private market3 but a classical (Aristotelian) distinction between the
public as an expression of the polis (where actors gather “to speak and act
together”) and the private as an expression of the oikos (Greek for household
and the root of the word economy) (where actors inhabit a domain of animal
necessity and are compelled to pursue their own interests for their survival).
For our purposes here, the oikos includes several institutional actors that usu-
ally are thought to be “public” yet that seek private interests for their own
survival: the Soviet military men, with state backing, wielded the threat of
nuclear destruction and personalized violence on the modern world; the
Party leaders pursued their own interests independent of the people; the eco-
nomic bureaucrats secured their own welfare apart from the welfare of the
economy; and the citizen workers tried make ends meet in their private lives.
The oikos, or the domain of the private, saturated the larger OGAS situation,
and the history of modern networks, including but not limited to Soviet
attempts, can be reread as a tale of private forces run amok.
These terms reframe our portrait of the challenges that were faced by
Soviet network projects. The problem was not that the state failed to regulate
private interests but that (according to Arendt) Marx put on a pedestal the

9800.indb 194 6/2/16 3:05 PM

Conclusion 195

speechless laborer (animal laborans), not the enlightened actor. The socialist
state served and scaled up the most private and basic of human needs but no
more. For Arendt, the equality of workers is tautological in the sense it equates
people on the basis of animal need, and the equality of citizens should be
sought by leveling unequal humans to create a better common world. She
also targeted elsewhere the teleological violence rendered by Hegelian his-
torical ideologies, such as Marxist-Leninist dialectical materialism: any state
convinced of its own historical path is sure to bring ruin to itself and others.4
In fact her critique of what she calls the rise of the social cannot be reduced
to the ruinous rise of socialism (whether Soviet, German national, or other
form) because her terms describe a range of modern advanced states that
have led the ongoing global scientific-technological revolution.
For the purposes of this book, the rise of state and market as parts of a
larger private household suggests the purpose of the command economy in
both theory and practice. In theory, it collapses private economic interests
into matters of state, and in practice, the state bureaucracy collapses into
the institutional turmoil of private actors. We can also see that Communist
Party leaders worked feverishly to secure their own power above all other
concerns and that the military shielded the Party, spied on foe and friend
alike, cannibalized resources, and separated itself from the national econ-
omy. The name of the Komitet Gosudarstvennoi Bezopasnosti (KGB) (Com-
mittee for State Safety) is similar to the name of the Committee of Public
Safety during the Reign of Terror in the French Revolution, except that the
Soviet version openly protected the state, not the public. The minister of
defense made a related point in 1965 when he rejected any collaboration
with the nascent OGAS Project. He identified the “healthy body” that his
military served not as the public but as “the government mechanism,” call-
ing the economic welfare of the nation “a scab.”5 In view of the divisions
in the Soviet oikos, this odd metaphor that the military was the mind of the
state body (not that the state was the mind of the economic body) appears
suddenly sensible.
Arendt’s concerns about the escalation of private interests over public
ones also explain why the OGAS story was not a people’s history and why
Glushkov addressed his last book to children, admitting that the workers
were not prepared for the OGAS. Soviet citizens lacked mechanisms for
mobilizing political will at scales larger than the dinner table, dacha, and
press editorial, so they had few chances to observe a public hearing of the
OGAS Project and far fewer chances to live a public life (or vita activa, as
Arendt fancied it). By rotating the private-public distinction from one
of market and state (and the state-market contradictions of the cold war

9800.indb 195 6/2/16 3:05 PM

196 Conclusion

economic order) to one of survival and political action, our vocabulary

maps onto more private divisions in the Soviet household. The problems
besetting the modern human world are far bigger than can be understood
from any particular pole of the cold war and may even be shared between
the two, as Arendt observed while she was in the middle of it.
Her argument comes with limitations. Like most political theory and
commentary, it offers no concrete proposals for reforming the current situ-
ation. It idealizes a polis of ancient Greece that did not exist. It also gave no
credit to the meaningful, vibrant, and even mischievous private lives that
Soviet citizens experienced in the workplace, such as the Cybertonia case
study (although Arendt notes that social gatherings that aggregate private
interests can be charming but never glorious, a fitting summary of almost
all virtual worlds and social media ever since). Moreover, her framing of
the rise of the social cannot be used to describe the asymmetric inequalities
of capitalism and social wealth because of the limitations of her founding
image of the oikos as rooted in the private household. That image of the
oikos would need to be subjected to a feminist philosophical critique of
the power inequalities that are buried in the history of the household and
domesticity—a critique that falls outside the scope of this book.6
To admit disillusionment with the normative values that organize
modern society is not necessarily to despair of the modern world itself,
which has brought with it extraordinary and positive advances. But it is an
attempt, like Sputnik, to glimpse new perspectives of the modern networked
world and then to rejoin the search for ways, like Soviet cybernetics, to har-
ness private power into the service of improving the human condition. A
few general comments on the modern world and its networks follow.

Contingency, Failure, Politics

Not only could our networked world have been otherwise—it can still be
otherwise today. One of the values of negative histories such as this one
is the reminder that most technological projects “fail” or never come to a
decisive end (perhaps both failure and repair occur in the long run).7 The
history of technology shows that most technological projects are not con-
sequential at all—at least in the conventional sense. Technological designs
are continuously not realized in operable material form and reproducible
prototypes, and the social processes that sustain scientific discovery rarely
arrive at a clear consensus. The historical record layers documentation of
the fossils and footnotes of “dead” media and their iterant afterlives.8

9800.indb 196 6/2/16 3:05 PM

Conclusion 197

This great apparent failure rate in innovations should help shape our con-
siderations of the causes and consequences of modern technologies, such as
computer networks. Contingent histories also help focus public debate better
than do popular histories of technology that parade about hackers, geniuses,
and geeks marching to the Whiggish beats of technological progress. In
negative histories, failures, even epic breakdowns, are normal. Astonishing
genius, imaginative foresight, and peerless technical wizardry are not enough
to change the world. This is one of the lessons of the OGAS experience. Its
story places the conventional concepts of technological successes and failures
on the wobbly foundations of the accidents of history. The historical record
is a cemetery overgrown in short-lived technological futures: stepping off its
beaten paths leads us to slow down and take stock before we rush to crown
the next generation of technologists as agents of change.
Perhaps the most hopeful reminder to would-be agents of social change
is also the hardest: the OGAS team understood that technological reform
is also political reform. A well-connected, talented team spent a genera-
tion fighting for the political life of a significant project—and those efforts
were not enough. Pity the scientists (and popular observers of science) who
believe that because we can isolate technical values in our minds, memos,
and mathematics, the alchemies of technological development will tri-
umph. Technologies are both artifacts and agents of change—a point that
has been made since Max Weber’s elective affinities (between Protestantism
and capitalism) and Ludwig Fleck’s social construction of science.9 In the
multivariable calculus of social reform, the only thing more certain than
the injunction that one must try to change the world (and media technolo-
gies are one among many ways to do so) is to admit there is no guarantee
that any given effort ever will.

A Nod toward Comparative Networks

How does the Soviet case compare to others? The OGAS tale intimates that
among the many variables in midcentury network projects—in this case,
Soviet socialism, cybernetic science, and decentralized networks—the most
important is the institutional environment for technological development.
Local institutional behavior is the concrete or quicksand into which the
history of networks is poured. Unlike the civilian-oriented Soviet OGAS
Project, the Chilean Cybersyn Project, and the (commercial) French Mini-
tel network, the military-initiated U.S. SAGE and ARPANET projects had
major effects on civilian industry and society. If there is a virtue to the post-
war American military-industrial-academic complex, perhaps it is that the

9800.indb 197 6/2/16 3:05 PM

198 Conclusion

complex allowed for cross-sector knowledge exchange and innovation trans-

fer. The failure of the Soviet knowledge base was arguably that the Soviet
military consumed resources and hoarded innovations from the civilian
Secondary to that argument, international communication networks pre-
cede international computer networks. Without international cybernetic sci-
ence discourse, the local dialects of systems science in the USSR, Chile, and
the United States could have taken different paths and perhaps found design
analogies for national networks other than the human mind (for example,
the socialist network as a nervous system in the body of the nation and the
liberal network as a neural network in the brain of the nation).10
The other huge socialist state anchoring the Eurasian steppe makes a
good comparison point. The People’s Republic of China is, like those states
in the former Soviet territories, a socialist state that is now devoted to devel-
oping mixed capitalist markets without democracy. Both China and Rus-
sia today operate according to informal networks of influence (guanxi and
blat) and are commercializing international computing innovations. The
sleek Baidu search, Youku video, and Sina Weibo microblogging platforms
imitate and improve the functionalities of Google search, YouTube video,
and Twitter. Both states also implement state controls to control national
computer network traffic. The most impressive of these is the “great firewall
of China,” which permits elites and technical experts an escape hatch from
the Chinese walled-garden version of the global Internet.
International communication networks also helped to jumpstart and
also consign to limbo local computer network projects. This account high-
lights three case studies: first, Anatoly Kitov’s discovery of Norbert Wiener’s
Cybernetics in a secret military library set into motion an internal transi-
tion in Soviet scientific discourse; second, Donald Davies and the British
Telecom industry prompted the U.S. government to revisit Paul Baran’s
RAND research on distributed packet-switching networks; and third, news
of the ARPANET going online in 1969 prompted the Politburo to revisit the
decade-old OGAS proposal in 1970. In each case, international communi-
cation networks (even when they were closed or secret) initially prompted
internal institutions to revisit concurrent innovations closer to home. As
it is in war, so it is in technology: rivals mimic each other mimicking each
other. Even so, cold war research networks were evidently too fixated on
the international exchange of knowledge among distant friend and foe.
Baran openly published his research in the early 1960s, for example, which
appears to have delayed his supervisors from attending to his work for
several years. Soviet scientists would have discovered Wiener’s Cybernetics

9800.indb 198 6/2/16 3:05 PM

Conclusion 199

years earlier, and likely to far less sweeping effect, had the book not been
banned (and had Stalin not repressed enemy sciences so vigorously). The
OGAS proposal probably would have received a fairer public hearing had
it not been a secret state project (and had there been a robust Soviet pub-
lic to share it with). Stretched between contending households that were
fueled by the same knowledge anxiety, cold war communication research
networks left few researchers with honor in their own lands. In cold war
science research, it appears that the more distant and closed the discovery,
the easier our narcissism; the closer and more open the discovery in states
of emergency, the easier our negligence.

Making Modern Network Culture Strange

The story of the OGAS Project reveals a network culture whose design val-
ues—the cybernetic nervous system of the nation, socialist technological
utopianism, and decentralized computer networks—now appear to be pecu-
liar to its own time and place. This sustained glance at the strangeness of
socialist network projects helps make familiar the foreignness of the mod-
ern network culture in historical relief. Consider a hardy perennial of new
media thought, the politics of technological utopia, for the OGAS Project
was nothing if not a projection of an intrepid socialist future. Socialist pol-
itics are no strangers to expansive, sometimes wild flights of imagination
about the bounteous blessings of technology. Although technological uto-
pianism belongs to social projects of all types, the socialist tradition boasts
a special breed of thinking, including the French socialist utopian thinker
Charles Fourier (whose early interests in architecture and engineering were
thwarted and who later worked briefly in Paris as head of the Office of Sta-
tistics), Karl Marx (who theorized about a socialist revolution near the end
of the Industrial Revolution in London), Nasser in Egypt, Tito in Yugoslavia,
Nehru in India, the Fabian Society and Labor Party in the United Kingdom,
Allende’s Cybersyn Project in Chile, and most recently the (independent)
Pirate Party of Sweden.11 In each of these cases, the socialist impulse seeks to
flatten out social relations, structurally reorganize society, automate and ease
labor, roll out statistical (state) accountability, and gather knowledge that
lightens, lifts, and liberates people (even though the effects of such techno-
logical utopianism often leans toward shades of dystopia).12 By imagining
the OGAS as a means to a brighter networked Communist future, its archi-
tects brought upon the project the full brunt of the oikos-led inequalities that
drove the administration of Soviet socialism. Perhaps the cardinal mistake
of the socialist imagination of technology is not to dream the celebrated

9800.indb 199 6/2/16 3:05 PM

200 Conclusion

dream of social justice but to bulldoze the rutted world of human relations
with the private interest logics of the oikos (military, corporations, states, and
individuals that seek only their own survival).
The Soviet OGAS figured out the “why?” (socialist utopia) but not the
“how?” for their large computer network projects, and researchers at the
U.S. ARPANET knew the “how?” (packet-switching networks) but not the
“why?” of modern networking. The Soviets’ missing “how?” lasted for the
duration of the project, and the absence of the Western “why?” remains
both its historical attraction and the contemporary challenge to computer
network culture.
The Western network “how?” has sped many unfinished attempts at
answering the network “why?” The technical openness of packet-switching
networks to diverse actors has afforded the Internet astonishing and well-
documented successes of technical energy, commercial innovation, and
cultural creativity. At the same time, the open-ended “why?” that has per-
mitted such generativity has also tolerated the entrance of private forces
that are interested in seizing possession of the operating systems and com-
munication infrastructures that mediate the globe. What Arendt observed
in the age of Sputnik still holds true in the age of smartphones: our tech-
nological capacity exceeds our political will to negotiate the terms of that
capacity. Our networks are no longer flat (if they ever were) but rather are a
consequence of network openness. Our lot, like that of the Soviets, is to live
in complex heterarchical power arrangements. Open network cultures are
slouching toward tethered devices, nonportable applications, walled gar-
dens (closed platforms), mobile contracts, and much else online and off. At
the individual level, these developments further feed and speed the parallel
encroachments of private communication forces worldwide, especially the
recently documented unprecedented surveillance of national and interna-
tional communication networks by governments and corporations in the
United States and the United Kingdom. Surveillance is the massification of
private attention and the antithesis of public attention (the first is a form
of global private labor, and the second, personal action).
Two generations ago, a few Soviet actors thought that the OGAS was a
good idea. Many more thought that it was a bad idea, and the many won
out. A generation ago, many Western observers thought that the Internet
was a good thing. The many this time were wrong. The Internet is not a
good thing, and it is not a bad thing. It is not a thing at all. The Internet is
many things, and many of those things are far less pleasant than cat videos
(cat videos feature creatures that, like many human spectators online, enjoy
the asocial separation that the screen affords them from their viewers).13

9800.indb 200 6/2/16 3:05 PM

Conclusion 201

This time, however, a few complex private forces are winning out, despite
the delusions of digital utopianism or quietism. Whatever else the Inter-
net is (interoperable, generative, nonproprietary, a platform for other plat-
forms), it is not public. As the history of the OGAS indicates, when the
public will to confront the high costs of modern network cultures is absent
or abused, private forces gladly rush in.
Consider the consequences of this Arendtian argument for modern
debate about publicity and privacy. It suggests one way of rereading the
term privacy in light of rise of the private logics of the oikos. Just as the Eng-
lish term publicity now belongs to the corporate practice of public relations,
so too does the term privacy (well before its legal coining as the right to be
left alone in 1890) belong to the private concerns of the state and the mar-
ket, not the person.14 In this sense modern privacy is not about the proper
spacing of the individual self and the other. It is about the sum of private
institutional interests that adjudicate the proper spacing of their institu-
tional homes (oikos) and the public. War rooms, closed sessions of the Sen-
ate, and boardrooms are where modern-day “privacy” resides, in the sense
that these are the institutions most interested in “the state of being privy
to” the lives of the public. Perhaps due to a mistaken understanding of pri-
vacy that emphasizes the individual, not the institution, scholars find the
term in “disarray” almost unintelligible outside a particular institutional
context, and other languages have trouble translating the English-language
lexeme. Perhaps we have misunderstood the term privacy all along.15 It is
not what Soviet citizens, under surveillance, never enjoyed. It is the rise of
the compulsive power of private forces themselves, which the USSR (among
other modern states) was permeated with. Private parties (including the
Party) and private secretaries (no matter how general or particular) directed
organizational forces (however informal, decentralized, and unpredictable)
that were bent on securing their own survival at the cost of others. The
term privacy has not been refeudalized so much as it stands for the colonial
expansion of the fiefdoms of institutional power.
Perhaps privacy scholarship should not seek to recover lost individual
privacy (the right to control the disclosure of personal information or alter-
nately the right to be left alone) but should critique the malignant growth
of institutional privacy (the right to own and its expansion to immor-
tal entities) whether the all-seeing eyes and ears of Google, the National
Security Agency, an OGAS-led command economy, or other institutions
engaged in massive amounts of information processing (in each example,
the economic liberal distinction between private corporation and public
state obscures more than it reveals). Glushkov’s computer networks would
have made the oikos of Soviet state-corporation even more privy to the

9800.indb 201 6/2/16 3:05 PM

202 Conclusion

work lives of the Soviet people. In a sense, this is precisely our lot now: the
networks that organize oikos powers are not hierarchical or decentralized
(like the institutions that check them). They are ambiguous, multiple, and
heterarchical. They vie for our attention, time, and action. In a time when
corporations roam the earth as legal persons, the shadows of Soviet net-
works are cast on the walls of the present. We might add to Adam Smith’s
famous warning that businessmen seldom meet without plotting against
their consumers: generals, politicians, and the clerisy fare not much bet-
ter when rolling out the privatizing logic of domination and need. With
few exceptions, large networked organizations are inclined to restrain each
other only when they interfere with one another in the common race to
privatize—or to use—the user. Since before Sputnik, our skies, screens, and
social lives have been filling with the drones of private network power.
Then and now, the polity and policy landscapes are not identical, and
we should not imagine them to be so. The private interests that kept com-
puter networks from being built in the USSR have since hijacked democratic
potentials in global networks. The basic institutions that stitch together
the social and political fabric of democratic society—the rule of law, func-
tioning courts, equitable tax compliance, Madisonian checks and balances,
human and civil rights, an independent press, and private institutions—
underlie the often ambiguous and always limited moral foundations of all
modern information societies and economies, even informal economies.16
The patronage socialism of the Soviet Union (like the crony capitalism of
modern-day Russia) was missing many of these elements (it had no rule of
law, no predictability of procedure, no regulated financial environment, no
bankruptcy law, no antirust law, no courts for managing property disputes,
and no virtuous regulation of inseparable market and state), but this rou-
tine criticism risks ignoring the bigger picture.
Perhaps the choice in the era of cybernetworks has never been between
the state and the market as the dominating metaphor for modern networks.
We need not accept as final either Glushkov and Cooley’s analogy of the
state as a nervous system (and the nation as its economic body) or McCull-
och and Baran’s analogy of the nation-state as a brain (and the network
as its neural net). Perhaps the way forward begins with criticizing both
cybernetic network analogies for privileging the image of the private mind
as supreme. The dominant metaphors for midcentury networked econo-
mies—market and state—move us no further than the cybernetic, and ulti-
mately human, hubris that the human mind organizes the world.
Although the landscape between the OGAS Project and the Internet
today varies widely, our hopes and despairs pivot on the same things that

9800.indb 202 6/2/16 3:05 PM

Conclusion 203

concerned the creators of the OGAS Project. There is a potential moral

authority in institutions and communities to check or caution the large
and unscrupulous actors that are intent on networking the world with the
creeping private logics of domination. We face new challenges, even as we
continue to target the cruelty, corruption, and compulsion of the world
that bears us—in every desire to limit the private mind of the oikos, there is
already a drop of our common human condition.
The OGAS story, therefore, is not only a tale that took place long ago
and far away. It can be seen as an allegory of our own lot today. The private
forces that were hard at work in the OGAS story are also hard at work in
the modern media environment. Informal networks abound, for better and
worse. We should not gaze at the OGAS Project from a comfortable distance
but realize how close its story hits to home. A world of difference separates
all allegories, but looking in the rearview mirror of history, the distance
between networked private powers is often closer than it appears.

Coda: A Contingent Legacy of Modest Networks

Beneath the modern imagination of smooth steel-brushed machines

interlinked by wires, signals, and smart protocols pulse the vibrant social
networks of relations whose virtues and vices have long been part of the
human condition. To understand modern networks is at root an exercise in
social self-discovery. Our network world shares with the fate of the OGAS
Project the vices of self-interest, apathy, back-stabbing, vain imaginations,
stupid conceit, poshlost’ (roughly the “self-satisfied vulgarity” of the petty
businessman and administrator, such as Chichikov in Gogol’s Dead Souls),
and all the rest. At the same time, it also shines brightly with generosity,
engagement, visionary insight, genius, byitie (another untranslatable Rus-
sian term meaning roughly “being,” “apperception,” or a higher state of
conscious reality that is resonant with Heideggerian being and scriptural
genesis), and much more. The networks binding the human condition can
be neither separated nor reconciled. Modern observers can no sooner state
the optimal conditions under which humankind has or will best enter the
age of global computer networks than we can solve the puzzle of the human
condition itself, although the attempts to solve the puzzle are worthwhile.
Given that there is no magic solution to these questions, we might do best
to seek a modest and cautious perspective on the causes and consequences
of the Soviet network experience.
Let us return for a moment to an earlier sense of the word technology.
In English usage until the early twentieth century, technology was not the

9800.indb 203 6/2/16 3:05 PM

204 Conclusion

hard stuff of tractors and circuits boards but was the study of industrial arts,
crafts, and techniques that organize, reveal, and frame the modern world.17
The suffix -ology in the term technology also appears in the term biology, the
study of life. Perhaps by understanding techne- as the artifacts of accultur-
ated human culture (behavior, gesture, oral, literate, print, industrial, mass,
and information media and much else), the term technology gives momen-
tum to the study of the crafts of social life.
The Soviet network history teaches several lessons. First, the ambitious,
far-seeing faith in the social consequences of technology is no guarantee
of technological change in the modern information age. Presentists who
look back at the science fiction, fact, and factions of Soviet cybernetics
may divine in these pages prophetic prefigurations of modern-day cloud
computing, e-commerce, big data processing, and much else. Opportunists
may be tempted to enthuse about recuperating the unrealized possibili-
ties of macroprocessing, natural language programming, a self-governing
economy, and perhaps even digital immortality, although they will do so
in their own tones and cadences. Second, Marx got the point of technol-
ogy wrong. He wrote that the relations of production—the social relations
that all people must enter into in modern life—are fundamental to all else.
Another lesson of the OGAS Project is that far more substantial than the
hard stuff of technology (cotton mills, industrial factories, hydroelectric
dams, nuclear power plants, and the factory and federated computer net-
works examined here) are the subtle, mundane techniques that continu-
ously work themselves out in the complex relations that constitute being
social. Finally, the critic Raymond Williams was right to attend to what
might be called the means of sociocultural production, not just the means
of industrial production. We can push the point further: the technological
means of world production are not just the mass media of newspaper, radio,
television, and computer but every commonplace device, understated tech-
nique, and learned skill—from a baby’s first vocalization to the experienced
insider’s knowledge of a bureaucracy’s peculiarities. These technologies and
techniques, creatively read, produce and manage a more genuine base for
understanding the arrangement of relations in modern society.18
That subtle and modest techniques hold sway over sophisticated infor-
mation technologies is a clear moral to this story. Letters to leadership found
their own random paths in the packet-switching labyrinth of Soviet state.
Everything—sudden success, interception and dismissals, evasive telephone
calls—came in reply. In fact, the first civilian-military national network pro-
posal anywhere was scuttled because a supervisor did not intercept one let-
ter but did intercept the next (such was the post in the Soviet military). The

9800.indb 204 6/2/16 3:05 PM

Conclusion 205

early institutional alliances between the Central Economic-Mathematical

Institute and the Institute of Cybernetics drifted apart over differences about
the scale (micro and macro, respectively) at which the mathematical tech-
niques for modeling economic relations should be carried out. The OGAS
Project—the ambitions of technocratic economic reform by network—was
nearly approved and funded except that two chairs at a committee meeting
went unoccupied. National technical networks connecting factories were
approved but never realized at the same time that local computer centers
in those factories were built but never interconnected—all because of coor-
dination problems (our coordination problems are as great today as their
solutions are subtle). Sophisticated chess algorithms outmaneuvered long-
term national planning methods and even the occasional chess master, but
never to the same effect as a simple notational system kept on index cards
(and now online databases). Ministerial ecosystems of paperwork collided
and proliferated, and the committee meeting—that omnipresent black box
of bureaucracies (even written minutes leave opaque the logics of small-
group decisions)—remains among the most undertheorized and delicate
techniques governing modern private power networks. Trains and tele-
phone calls were taken and missed; doors opened and locked; hearts and
minds pushed to their limits—and sometimes beyond.
The history of Soviet networks showcases something more enduring,
powerful, and subtle than a plumbing and sounding out of the stately
heights of electronic socialism (although it also does that). It reveals the
modest media on which our social relations turn—labyrinthine commit-
tee reports and paper trails, bureaucratic and budgetary categories that
scrimmage careers, the semantic vagaries of public press releases and pre-
cise accounting, empty chairs and scattered letters, accidental meetings in
hallways and dachas, and all the other errata of the constant communica-
tion and infrequent communion that arrange our lives. When I set out to
research the Soviet networks, I hoped for historical insights into the media
of tomorrow, but what I found instead were dusty, derelict, and sometimes
dispensable residual artifacts of a technological vision for a labyrinthine
state now largely forgotten. Not only was I wrong to look for a peek into
the future in the archives of the past, I was wrong to think I had not found
them. Because the techniques of paper knowledge and print culture con-
tinue to accumulate in the scattered anecdotes and artifacts that make up
our societies and the stories we tell about them, they too will likely endure
as the media of tomorrow. These are the media technologies, writ large, that
govern the computer networks and other props of the current information
age; theirs are the modern media networks that matter most.

9800.indb 205 6/2/16 3:05 PM

206 Conclusion

The OGAS Project, like most information age projects, has more of
bureaucracy than bits to it. The history and perhaps the future of the cur-
rent information age will have less to do with the next generation of futur-
ist technologies than it will with the networks of actors and institutions
governing the conditions of social relations and the use of knowledge. It
would be a mistake to conclude that this far-seeing generation of Soviet
scientists and technologists did not realize a network that was capable of
changing the world. Their dreams and ambitions were realized not in the
networks of steel and silicon chips but in the networks that long have and
will continue to govern our lives. The All-State Automated System Project
lives on as a story refracted in the records of print culture. In the end, the
story told here tells its own moral and method. It asks us to distinguish
and extract it from the swirling and glorious strangeness of all scientific
ambition that buoys the modern world, exert good will to tolerate it in its
oddities, critique it not for what it has not accomplished but for its courting
of the irresistible enchantment of modern-day network visions, and finally
perhaps even to grow used to it, to wait for it, and to have one day admitted
its passage and place into the greater living network of ideas and institu-
tions that make up the modern world.
Such is the uneasy history of Soviet networks. Networks are not the appli-
cation of a theory of networks, nor are they the children of hard gadgetry
and pragmatic engineering. They are the technical arrangements of social
relations that have and will continue to change the world. Much remains
appropriately and implicitly contingent and unpredictable in the historic
making and unmaking of global networks. May the story of the Soviet net-
works and their troubled paths into an alternative information age stand as
sentinel cautions for our networked times. It is not in the nature of daring
ideas and the routines of history to come to an end, although such is the
lot of books.

9800.indb 206 6/2/16 3:05 PM


A c

No book can exist without supporting scholarly and institutional networks,

and this is especially true of this book, given its profound debt to men-
tors, colleagues, friends, and institutions that I engaged with along the
A c
way. Whatever commonsense and clear writing is contained in this book—
which is an expansion and complete reworking of about half of my Ph.D.
dissertation (Columbia University, 2010)—is due to the patient influence
of Michael Schudson, a model mentor, adviser, and scholar. My research
interests owe Todd Gitlin, Richard John, the late Catharine Nepomnyash-
chy, and Siva Vaidhyanathan far more than their mentorship. In addition
to the four mentors at four schools to whom this book is dedicated (Gary
Browning at Brigham Young University, Fred Turner at Stanford University,
Michael Schudson at Columbia University, and Joli Jensen at the Univer-
sity of Tulsa), I have to acknowledge my debt to mentors and teachers such
as Craig Calhoun, the late James W. Carey, Monika Greenleaf, and Andie
Tucher, among many others. At the University of Tulsa, my colleagues Mark
Brewin, John Coward, and Joli Jensen have fashioned a work environment
collegial enough to make any young professor enviable.
A portion of the first chapter was previously published as the article
“Normalizing Soviet Cybernetics” and appears here with permission from
the journal Information & Culture.
I thank the good folk at my home Ph.D. degree in communications
program at the Columbia University Graduate School of Journalism as
well as the Archives of the Russian Academy of Sciences, the Institute of
Cybernetics of the Ukrainian Academy of Sciences, the Central Economic-
Mathematical Institute of the Russian Academy of Sciences, the Institute
of Cybernetics of the Ukrainian Academy of Sciences, Elena Vartanova and
the School of Journalism at Moscow State University for hosting research
visits, the Institute Archives and Special Collections at the Massachusetts

9800.indb 207 6/2/16 3:05 PM

208 Acknowledgments

Institute of Technology, and the Freedom of Information Act governing the

Central Intelligence Agency and the Federal Bureau of Investigation. I also
thank the patient staffs at Butler Library at Columbia University, Widener
Library at Harvard University, the Harold B. Lee Library at Brigham Young
University, and that modern-day library of Alexandria, interlibrary loan
and online scholarly databases. Support and fellowship have come from
the Oklahoma Center for the Humanities and faculty summer develop-
ment grants from the University of Tulsa, the Nevzlin Center for the Study
of Russian Jewry and the Lady David Postdoctoral Fellowship at Hebrew
University, the Kenneth E. and Becky H. Johnson Foundation, and Junior
and other fellowships from the Harriman Institute at Columbia University.
Audiences have contributed much at Yale Law School, the Columbia Uni-
versity School of Journalism and the Harriman Institute, the University of
Tampere, the Mohyla School of Journalism, the Princeton University Cen-
ter for Information Technology Policy, and many others. The extraordinary
communities in the orbit of Jack Balkin’s Information Society Project at Yale
Law School and the Berkman Center for Internet and Society at Harvard
University have delighted and engaged me for nearly a decade now.
The research animating this book could not exist without the base of
work laid by Slava Gerovitch: I am grateful to him for his mentorship.
Without the friendship and historical scholarship of Aleksei Viktorovich
Kuteinikov and especially the grace and resourcefulness of Vera Viktorevna
Glushkova, this introduction to Viktor Glushkov’s story to the English-
speaking world would likely not exist in book form. I am also grateful to
Vladimir Anatolevich Kitov for helpful scholarly resources about his father,
Anatoly Kitov, as well. I thank them all three. Previous drafts have benefited
from the valuable comments of Geof Bowker, Peter Sachs Collopy, Paul
Edwards, Bernard Geoghegan, Lydia Liu, Eden Medina, and Mara Mills on
cybernetics and information theory, while Alex Bochannek, Elena Doshly-
gina, Michael Gordin, Loren Graham, Martin Kragh, Adam Leeds, Ksenia
Tatarchenko, and others have taught me much about the Soviet situation.
At the risk of leaving many others unnamed, I would also like to thank
Colin Agur, Karina Alexanyan, Chris W. Anderson, Mark Andrejevic, Rose-
mary Avance, Burcu Baykurt, Valerie Belair-Gagnon, Jonah Bossewitch,
Gabriella Coleman, Laura DeNardis, Jeffrey Drouin, Maxwell Foxman, Alex-
ander Galloway, Gina Giotta, Abe Gong, Eugene Gorny, Orit Halpern, Lewis
Hyde, Andryi Ishchenko, Carolyn Kane, John Kelly, Beth Knobel, Liel Liebo-
vitz, Deborah Lubken, Kembrew McLeod, David Park, Ri Pierce-Grove, Amit
Pinchevski, Jefferson Pooley, Erica Robles, Natalia Roudakova, Chris Russil,
Jonathan Saunders, Limor Schifman, Trebor Scholz, Steven Schrag, Zohar

9800.indb 208 6/2/16 3:05 PM

Acknowledgments 209

Sella, Lea Shaver, Bernhard Siegert, Peter Simonson, Thomas Streeter, Ted
Striphas, Patrik Svensson, McKenzie Wark, David Weinberger, and Jonathan
Zittrain for helpful conversations and comments on drafts over the years.
To me goes the real award of association with them. Menahem Blondheim,
Paul Frosh, Elihu and Ruth Katz, Amit Pinchevski, and Limor Shifman,
among others, gave me far more than a year of intellectual stimulation and
companionship at Hebrew University; Lucas Graves, Rasmus Nielsen, and
Julia Sonnevend make ideal conversation partners in our ongoing search
for social theoretical understanding. Only two people, to my knowledge,
have read the whole manuscript—John Durham Peters and Audra Wolfe.
To my delight, both have provided invaluable professional criticism as both
scholars and editors; the book is better for the portion of their comments I
have incorporated and diminished by those left undone. My editor, Sandra
Braman, has buoyed my writing process with her abiding support of the
project from start to finish. These are some of the people—alongside the
good folk at MIT Press, especially Margy Avery, Deborah Cantor-Adams,
and Rosemary Winfield as well as the generous and detailed criticism of two
anonymous reviewers—that have improved the book in your hands.
Scholarship began for me first as a family affair. Anyone who knows
the depth and breadth of my father’s intellectual generosity—a way of life
and a man I have long looked up to—should also know that I am at home
in my father’s field because I am my mother’s son. My enduring love and
gratitude go to Marsha Paulsen and John Durham Peters for modeling what
matters. My own family has made considerable sacrifice and contributions
as well. My favorite ruffians and readers—Aaron, Elliot, Libbie, and Maya—
have made precious the hours I have spent working on what Libbie has too
generously titled “a very long book by Tato Peters” and more precious still
the balance of time I have with them. Above all, Kourtney Lambert—my
love and favorite reader—makes it all worthwhile.

9800.indb 209 6/2/16 3:05 PM

9800.indb 210 6/2/16 3:05 PM



9800.indb 211 6/2/16 3:05 PM

9800.indb 212 6/2/16 3:05 PM
A  Basic Structure of the Soviet Government

This brief appendix provides a simple outline of the complex and chang-
ing structure of the government of the Union of Soviet Socialist Repub-
lics (USSR). The country was divided into one federated socialist republic
(Russian) and fourteen soviet socialist republics (Armenian, Azerbaijan,
Byelorussian, Estonian, Georgian, Kazakh, Kirghiz, Latvian, Lithuanian,
Moldavian, Tajik, Turkmen, Ukrainian, and Uzbek). Each republic con-
tained stacked (and sometimes confused) subdivisions ranging from small-
est to largest in this order: raion (districts, areas, subdistricts), krai (territory),
okrug (district), and oblast’ (region).
The basic structure of the Soviet state had three parts or political bod-
ies—the Communist Party, the bureaucracy, and the legislature (this
ignores the mostly toothless judiciary of the Supreme Court). The Com-
munist Party of the Soviet Union, the only political party permitted by the
constitution, coordinated all the affairs of the economy and society. The
pyramid party structure rested on a selection of Soviet citizens (no more
than 9 percent of the Soviet people were ever members of the Communist
Party), and membership was overwhelmingly made up of professional and
often technocratic males (the Party shares this with the current digerati
demographic). The party structure stretched upward from the members to
local party organizations, to local, district, and regional congresses, to the
National Party Congress, to the Central Committee, and finally to the Polit-
buro, which was the governing Party committee of the land. At the head
of the Politburo sat—in a fitting encapsulation of the Party’s bureaucratic
spirit—the general secretary, a position that Stalin granted almost supreme
powers after Lenin’s death. The general secretary worked in theory along-
side the premier (the bureaucracy) and the president of state (the legisla-
ture) and oversaw the Secretariat, a second ruling Party committee on a
level with the Politburo.

9800.indb 213 6/2/16 3:05 PM

214  Appendix A

The central structure of the bureaucracy scans simply but proved laby-
rinthine in practice. At the bottom again were the people, and at the top
was the premier, who oversaw the Council of Ministers. Between the citi-
zens and the Council of Ministers fell the internal structures of between
twelve and thirty-seven ministries (such as the Ministry of Agriculture and
Food and the Ministry of Transport Construction) and the military (the
Red Army). During economic reforms, ministries were regularly reorga-
nized, consolidated, and strengthened, and many of them worked across
local, district, and national committee subdivisions. This analysis under-
scores the Soviet bureaucratic divide between civilian ministries and the
military (which was a training ground for Party leadership and a sink for
the national budget).
Lastly, the legislature was constitutionally appointed in 1918 to over-
see economic, social, and security affairs, although in the latter half of the
twentieth century its power was largely secondary to the Party and the
bureaucracy. Citizen-elected local, district, and regional soviets (or coun-
cils) informed the Supreme Soviet, the Presidium, and the president or head
of state, whose powers paled in comparison to the premier (head of the
bureaucracy) and the general secretary (head of the Party). The Presidium
was initially a decision-making body that was a peer with the Council of
Ministers (bureaucracy) and the Politburo and Secretariat (Party), although
its influence waned with the consolidation and decentralization of power
in the Party and bureaucracies under and after Stalin.
These three branches of government were staffed by the nomenklatura
or elite responsible for higher positions of authority. Formally, the nomen-
klatura occupied a small, elite subset of the already elite Party membership,
although in practice it also could include the intelligentsia or needed experts
who did not have to be Party members (most of the scientists and adminis-
trators featured here were members of the Party and often the intelligentsia).
In the management of the command economy, Party and state hierarchies
were separate and overlapping. So although members of the nomenklatura
could manage a state-owned factory, they also had to have party approval if
they were not party members. In such cases, factory directors might report
to the local Party secretary as an ordinary Party member, and the Party sec-
retary would report to the director as an employee. In all, this book offers
a reminder that in the management of large organizations, especially the
Soviet state and economy, the questions of structure and governance are
rarely so straightforward as they may appear on paper.

9800.indb 214 6/2/16 3:05 PM

B  Annotated List of Slavic Names


For the ease of the English reader, the text refers to people who recur in this
history by first and last names; other persons, no matter how significant,
whose names do not appear in the text frequently are named in the Soviet
academic tradition of two initials (the first name and patronymic) followed
by last name. Only recurring figures are listed below.

Aksel Berg (1893–1973): Engineer admiral, deputy chair of the Council on


Mikhail Botvinnik (1911–1995): Soviet international grandmaster, founding

member of the Soviet school of chess, professional electrical engineer, com-
puter scientist, and champion of early computer chess Pioneer program,
and author of several proposals to computerize strategic planning.

Leonid Brezhnev (1906–1982): General secretary of the Union of Soviet

Socialist Republics (1964–1982).

Nikolai Fedorenko (1917–2006): Chemist and economist, director of the

Central Economic-Mathematical Institute (1963–1985), coauthor of the
EGSVTs (Unified State Network of Computing Centers) network project
(1963), academician.

Vasily Garbuzov (1911–1985): Minister of finances (1965–1980), principal

opponent to the OGAS (All-State Automated System) Project, rival of Vladi-
mir Starovsky and the Central Statistical Administration.

Viktor Glushkov (1923–1982): Prominent Soviet cyberneticist, director

of the Institute for Cybernetics in Kiev, Ukraine (1967–1982), author of
OGAS (All-State Automated System) (1963–1982), coauthor of the EGSVTs
(Unified State Network of Computing Centers) (1963) network projects,

9800.indb 215 6/2/16 3:05 PM

216  Appendix B

Mikhail Gorbachev (1931–): General secretary, Union of Soviet Socialist

Republics (1985–1991).

Leonid Kantorovich (1912–1986): Soviet economic mathematician, pioneer

in linear modeling, Nobel Prize in economics (1975).

Mstislav Keldysh (1911–1978): Mathematician, Soviet space theorist, chair

Soviet Academy of Sciences (1961–1975) (where he helped rehabilitate
cybernetics and genetics).

Aleksandr Kharkevich (1904–1965): Communication engineer, director of

the Institute for Information Transmission Problems (1962–1965), author
of the ESS (Unified Communication System) network project (1963).

Nikita Khrushchev (1894–1971): First (general) secretary of the Union of

Soviet Socialist Republics (1953–1964).

Anatoly Kitov (1920–2005): Mathematician, colonel engineer, first Soviet

cyberneticist, coauthor The Basic Features of Cybernetics (1955), author of
the EASU (Economic Automatic Management System) network proposal

Ernst Kolman (1892–1979): Failed mathematician, philosopher-critic,

accuser of Andrei Kolmogorov (1939), author of “What Is Cybernetics?”
(1955), first ideological supporter of Soviet cybernetics (1955–1979).

Andrei Kolmogorov (1903–1987): Prominent mathematician, public cyber-

netics supporter (1960–1970).

Aleksei Kosygin (1904–1980): Premier of the Union of Soviet Socialist

Republics (1964–1980), deputy chair of the Soviet Council of Ministers,
appointed Viktor Glushkov and Nikolai Fedorenko to develop the OGAS
Project and the EGSVTs (Unified State Network of Computing Centers) net-
work project (1962).

Aleksei Lyuapunov (1911–1973): Mathematician, pioneering cyberneticist,

coauthor of “Basic Features of Cybernetics” (1955).

Vasily Nemchinov (1894–1964): Economic mathematician, organizer of the

laboratory in Novosibirsk (1958) that became Nikolai Fedorenko’s Central
Economic-Mathematical Institute in Moscow (1963).

Konstantin Rudnev (1911–1980): Author of a 1963 Izvestia article in favor of

using computers in national planning, head of the Ministry of Instrument
Making, Automated Equipment, and Control Systems (1965–1980).

9800.indb 216 6/2/16 3:05 PM

Annotated List of Slavic Names  217

Sergei Sobolev (1908–1989): Prominent mathematician, coauthor of “The

Basic Features of Cybernetics” (1955), public supporter of cybernetics

Vladimir Starovsky (1905–1975): Director of the Central Statistical Adminis-

tration in the Council of Ministers (1957–1975), principal opponent of the
OGAS (All-State Automated System) Project, rival of Vazily Garbuzov and
his Ministry of Finance.

9800.indb 217 6/2/16 3:05 PM

9800.indb 218 6/2/16 3:05 PM
C  Network and Other Project Acronyms


ARPANET Advanced Research Projects Agency Network, later Defense

Advanced Research Projects Agency Network (1969–1983), United States
Department of Defense, first packet-switching network and predecessor to
the Internet.

ASU (avtomatizirovannaya sistema upravleniya): Automated system of man-

agement. The Soviet term for a management information and control
system, or, effectively, a local network between an onsite computer and
attending industrial processes that it supervises at a factory.

CEMI (tsentralnyi ekonomicheskii-mathematicheskii institute): The Central

Economic-Mathematical Institute of the Academy of Sciences in Moscow,
proposed by Nemchinov, built on his Laboratory of Economic Mathemati-
cal Methods, founded in 1963, first directed by Nikolai Fedorenko, and an
early collaborator with Viktor Glushkov’s Institute of Cybernetics on the
OGAS (All-State Automated System) Project and the EGSVTs (Unified State
Network of Computing Centers) network projects.

CSA (tsentral’noe statisticheskoe upravleniye): The Central Statistical Adminis-

tration (or Directorate) was, as part of the Council of Ministers (the highest
executive council in the Soviet Union) between 1948 and 1987, the main
organization in the Soviet state charged with statistical oversight.
EASU (ekonomicheskaya avtomatizirovannaya sistema upravleniya): Economic
Automated Management System proposed by Anatoly Kitov (1959).

ESS (edinaya sistema svyazi): Unified Communication System, a compre-

hensive data communication network planned by Aleksandr Kharkevich

EGSVTs (edinogosudarstvennaya set’ vyichisletel’nikh tsentrov): Unified State

Network of Computing Centers, technical base of the OGAS (All-State

9800.indb 219 6/2/16 3:05 PM

220  Appendix C

Automated System) Project, coauthored by Viktor Glushkov and Nikolai

Fedorenko (1963). In other literature, associated with a complex series of
other subdevelopments. The EGSVTs was a subset of the overall OGAS

OGAS(U) (obshche-gosudarstvennaya avtomatizirovannay sistema upravleniya):

All-State Automated System (of Management). Inspired by Anatoly Kitov’s
EASU (Economic Automatic Management System) and composed of a
national network connecting and managing ASUs (automated system of
management), it was proposed by Viktor Glushkov and others between
1963 and 1985, developed variously by the Institute of Cybernetics, CEMI
(Central Economic-Mathematical Institute), and others. EGSVTs (Unified
State Network of Computing Centers) was projected to be its the technical
base. SOFE (System for the Optimal Functioning of the Economy) was pro-
jected to be its modeling system.

SAGE: Semi-Automatic Ground Environment, an air defense control system

used by the United States and Canada from the late 1950s through the
1980s. Although ineffectual as a strategic network, it appears to have been
an important site for developing online, real-time interactive computing
over long distances.

SOFE (sistema optimal’nogo funktsionirovaniya ekonomiki): System for the

Optimal Functioning of the Economy, developed under Nikolai Fedorenko
at CEMI (Central Economic-Mathematical Institute), which pioneered
systems models and theories for optimizing economic planning since the
1960s. Initially a companion program for developing the optimization and
economic management software behind the OGAS (All-State Automated
System) Project.

9800.indb 220 6/2/16 3:05 PM




1. Flo Conway and Jim Siegelman, Dark Hero of the Information Age: In Search of Nor-
bert Wiener, the Father of Cybernetics (New York: Basic Books, 2005), 392 n. 318.

2. Marshall McLuhan, Understanding Media: The Extensions of Man (New York:

McGraw-Hill, 1964).

3. Slava Gerovitch, “InterNyet: Why the Soviet Union Did Not Build a Nationwide
Computer Network,” History and Technology 24 (4) (December 2008): 335–350.

4. Viktor Shklovsky, “Art as Technique” (1917), in Russian Formalist Criticism: Four

Essays, ed. Lee T. Lemon and Marion J. Reiss (Lincoln: University of Nebraska Press,
1965), 3–24.

5. Peter Brown, The Body and Society: Men, Women, and Sexual Renunciation in Early
Christianity (New York: Columbia University Press, 1988), xvii.


1. On September 19, 1990, fifteen months before the Soviet Union collapsed, the
Internet Corporation for Assigned Names and Numbers (ICANN) assigned the .su
country code top-level domain, and it remains in use today.

2. For more on Akademgorodok, see Paul R. Josephson, New Atlantis Revisited:

Akademgorodok, the Siberian City of Science (Princeton: Princeton University Press,

3. The literature on the Soviet Union’s role in the cold war is enormous. Readers
unacquainted with that literature may wish to start with a primer on the global cold
war context, such as Robert J. McMahon, The Cold War: A Very Short Introduction
(New York: Oxford University Press, 2003), Steven Lovell, The Soviet Union: A Very
Short Introduction (New York: Oxford University Press, 2009), and a more substantial

9800.indb 221 6/2/16 3:05 PM

222  Notes to Introduction

work by Orlando Figes, Revolutionary Russia, 1891–1991: A History (New York: Metro-
politan Books, 2014). Other classics outside the Soviet period or space include Eric
Hobsbawm, The Age of Extremes: A History of the World, 1914–1991 (New York: Pan-
theon Books, 1994); Orlando Figes, Natasha’s Dance: A Cultural History of Russia
(New York: Picador, 2003); and James H. Billington, The Icon and the Axe: An Interpre-
tive History of Russian Culture (New York: Vintage, 1966). For more on the intellectual
context, see the politically opposing pair, Isaiah Berlin, Russian Thinkers (New York:
Penguin Group, 1978), and Richard Pipes, Russian Conservatism and Its Critics: A
Study in Political Culture (New Haven: Yale University Press, 2005).

4. Robert E. Kohler and Kathryn M. Olesko, “Introduction: Clio Meets Science: The
Challenges of History,” Osiris 27 (1) (2012): 4–6.

5. The literature on the history of computing in the United States context is also
significant. For a basic introduction, see Paul E. Ceruzzi, Computing: A Concise History
(Cambridge: MIT Press, 2012); Paul E. Ceruzzi, A History of Modern Computing (Cam-
bridge: MIT Press, 1998); Martin Campbell-Kelly and William Aspray, Computer: A
History of the Information Machine (Boulder, CO: Westview, 2004); and William
Aspray and Paul E. Ceruzzi, The Internet and American Business (Cambridge: MIT
Press, 2008). The growing literature on the U.S. history of the Internet includes
works such as Janet Abbate, Inventing the Internet (Cambridge: MIT Press, 1999); Paul
N. Edwards, The Closed World: Computers and the Politics of Discourse in Cold War
America (Cambridge: MIT Press, 1996); Finn Burton, Spam: A Shadow History of the
Internet (Cambridge: MIT Press, 2013); and Thomas Streeter, The Net Effect: Romanti-
cism, Capitalism, and the Internet (New York: New York University Press, 2011). See
also Jonathan Zittrain, The Future of the Internet, and How to Stop It (New Haven: Yale
University Press, 2008), and Tim Wu, The Master Switch: The Rise and Fall of Informa-
tion Empires (New York: Atlantic Books, 2010).

6. Scholarship has not yet advanced a deep understanding of the relationship

between social justice and computing, although initial inroads are being made in
the critical study of gender and computing. A few works of note include Donna
Haraway, Simians, Cyborgs and Women: The Reinvention of Nature (New York: Rout-
ledge, 1991); Jennifer S. Light, “When Computers Were Women,” Technology and
Culture 40 (3) (1999): 455–483; Nathan Ensmenger, The Computer Boys Take Over:
Computers, Programmers, and the Politics of Technical Expertise (Cambridge: MIT Press,
2010); and Mette Bryld and Nina Lykke, Cosmodolphins: Feminist Cultural Studies of
Technology, Animals and the Sacred (New York: Zed Books, 2000).

7. David E. Hoffmann, The Dead Hand: The Untold Story of the Cold War Arms Race
and Its Dangerous Legacy (New York: Random House, 2009), 150–154, 364–369, 422–
423, 477.

8. Ibid., 153–154.

9800.indb 222 6/2/16 3:05 PM

Notes to Introduction  223

9. For sample references, see Kevin Kelly, Out of Control: The New Biology of Machines,
Social Systems, and the Economic World, Fourth Edition (Reading, MA: Addison Wesley,
2004), chap. 4; Eric Raymond, The Cathedral and the Bazaar: Musings on Linux and
Open Source by an Accidental Revolutionary (New York: O’Reilly, 1999); and Leon
Trotsky, Platform of the Joint Opposition (1927) (London: New Park Publications,
1973), especially “The Agrarian Question and Social Construction.”

10. Manuel Castells, End of the Millennium: The Information Age—Economy, Society,
and Culture (Malden, MA: Blackwell, 1998), 5–68; Lawrence Lessig, Code and Other
Laws of Cyberspace (New York: Basic Books, 1999), 3–8.

11. Yochai Benkler, The Wealth of Networks: How Social Production Transforms Markets
and Freedom (New Haven: Yale University Press, 2006).

12. Melvin Kranzberg, “Technology and History: ‘Kranzberg’s Laws,’” Technology and
Culture 27 (3) (1986): 544–560.

13. For Latour’s aphorism, see Bruno Latour, “Technology Is Society Made Durable,”
in A Sociology of Monsters: Essays on Power, Technology and Domination, ed. John Law,
Sociological Review Monograph No. 38 (London: Routledge, 1991), 103–132. For an
excellent bibliographical bridge between science and technology studies (STS) and
the study of information technologies, see P. Boczkowski and L. Lievrouw, “Bridging
STS and Communication Studies: Scholarship on Media and Information Technolo-
gies,” in The Handbook of Science and Technology Studies, ed. E. Hackett, O. Amster-
damska, M. Lynch, and J. Wajcman, 3rd ed. (Cambridge: MIT Press, 2007), 949–977.

14. Geoffrey C. Bowker and Leigh Starr, Sorting Things Out: Classification and Its Con-
sequences (Cambridge: MIT Press, 1999), 33–50.

15. Eric Hobsbawm, How to Change the World: Reflections on Marx and Marxism (New
Haven: Yale University Press, 2011), 22–41.

16. The article that made this book possible is Slava Gerovitch, “InterNyet: Why the
Soviet Union Did Not Build a Nationwide Computer Network,” History and Technol-
ogy, 24 (4) (2008): 335–350. See also Slava Gerovitch, “The Cybernetics Scare and the
Origins of the Internet,” Baltic Worlds 2 (1) (2009): 32–38; Slava Gerovitch, From
Newspeak to Cyberspeak: A History of Soviet Cybernetics (Cambridge: MIT Press, 2002);
Slava Gerovitch, “Speaking Cybernetically: The Soviet Remaking of an American Sci-
ence,” Ph.D. diss., Program in Science, Technology and Society, Massachusetts Insti-
tute of Technology, 1999; Loren R. Graham, Science, Philosophy, and Human Behavior
in the Soviet Union (New York: Columbia University, 1987); Loren R. Graham, Science
in Russia and the Soviet Union: A Short History (New York: Cambridge University Press,
1993); and Loren R. Graham, Lonely Ideas: Can Russia Compete? (Cambridge: MIT
Press, 2013).

17. Classic and recent histories of the Internet and its American milieu include
Abbate, Inventing the Internet; Edwards, The Closed World; Burton, Spam; and Thomas

9800.indb 223 6/2/16 3:05 PM

224  Notes to Introduction

Streeter, The Net Effect: Romanticism, Capitalism, and the Internet (New York: New
York University Press, 2011). For more popular introductions, see Ian F. McNeely
with Lisa Wolverton, Reinventing Knowledge: From Alexandria to the Internet (New
York: Norton, 2008), whose scholarly breadth and snap counterweight popular
accounts such as Katie Hafner, Where Wizards Stay Up Late: The Origins of the Internet
(New York: Simon & Schuster, 1996), and Walter Isaacson, The Innovators: How a
Group of Hackers, Geniuses, and Geeks Created the Digital Revolution (New York: Simon
& Schuster, 2014).

18. I owe a version of this line and much else to conversations with Elihu Katz at the
Department of Communication at Hebrew University in the spring of 2011.

19. Eden Medina, Cybernetic Revolutionaries: Technology and Politics in Allende’s Chile
(Cambridge: MIT Press, 2011).

20. The literature on cybernetics, viewed in its breadth, is considerable and growing.
For a brief introduction, see Bernard Geoghegan and Benjamin Peters, “Cybernet-
ics,” in The John Hopkins Guide to Digital Media, ed. Marie-Laure Ryan, Lori Emerson,
and Benjamin J. Robertson (Baltimore: John Hopkins University Press, 2014), 109–
112. For more on cybernetics in the United States, see Peter Galison, “The Ontology
of the Enemy: Norbert Wiener and the Cybernetic Vision,” Critical Inquiry 21 (1)
(1994): 228–266; Geoffrey C. Bowker, “How to Be Universal: Some Cybernetic Strate-
gies, 1943–1970,” Social Studies of Science 23 (1993): 107–127; Geoffrey Bowker, “The
Empty Archive: Cybernetics and the 1960s,” in Memory Practices in the Sciences (Cam-
bridge: MIT Press, 2006); Lily E. Kay, “Cybernetics, Information, Life: The Emergence
of Scriptural Representations of Heredity,” Configurations 5 (1) (1997): 23–91.Books
on the cybernetic context before and during the U.S. cold war include Edwards, The
Closed World; David Mindell, Between Human and Machine: Feedback, Control, and
Computing before Cybernetics (Baltimore: John Hopkins Press, 2002); Jennifer Light,
From Warfare to Welfare: Defense Intellectuals and Urban Problems in Cold War America
(Baltimore: Johns Hopkins University Press, 2003); and Darren Tofts, Annemarie
Jonson, and Alessio Cavallaro, eds., Prefiguring Cyberculture: An Intellectual History
(Cambridge: MIT Press, 2002).A few biographical works include Steve J. Heims, The
Cybernetics Group (Cambridge: MIT Press, 1991); Steve J. Heims, John von Neumann
and Norbert Wiener: From Mathematics to the Technologies of Life and Death (Cam-
bridge: MIT Press, 1982); Pesi R. Masani, Norbert Wiener, 1894–1964 (Boston:
Birkhäuser Verlag, 1990); Flow Conway and Jim Siegelman, Dark Hero of the Informa-
tion Age: In Search of Norbert Wiener, the Father of Cybernetics (New York: Basic Books,
2005); and Hunter Crowther-Heyck, Herbert A. Simon: The Bounds of Reason in Modern
America (Baltimore: Johns Hopkins University Press, 2005).A few key theorizations
and historical treatments include N. Katherine Hayles, How We Became Posthuman:
Virtual Bodies in Cybernetics, Literature, and Informatics (Chicago: University of Chi-
cago Press, 1999); Jean-Pierre Dupuy, The Mechanization of the Mind: The Origins of
Cognitive Science, trans. M. B. DeBevoise (Princeton: Princeton University Press, 2000;
Cambridge: MIT Press, 2009); John Johnston, The Allure of Machinic Life: Cybernetics,

9800.indb 224 6/2/16 3:05 PM

Notes to Introduction  225

Artificial Life, and the New AI (Cambridge: MIT Press, 2008); Philip Mirowski, Machine
Dreams: Economics Becomes a Cyborg Science (New York: Cambridge University Press,
2001); Orit Halpern, “Dreams for Our Perceptual Present: Archives, Interfaces, and
Networks in Cybernetics,” Configurations 13 (2007): 283–319; Stuart Umpleby, “A
History of the Cybernetics Movement in the United States,” Journal of the Washing-
ton Academy of Sciences 91 (2005): 54–66; Bernard Geoghegan, “The Historiographic
Conceptualization of Information: A Critical Survey,” IEEE Annals of the History of
Computing 30 (2008): 66–81.For more on cybernetics in the Soviet Union, see Slava
Gerovitch, From Newspeak to Cyberspeak: A History of Soviet Cybernetics (Cambridge:
MIT Press, 2002); David Holloway, “Innovation in Science: The Case of Cybernetics
in the Soviet Union,” Science Studies 4 (1974): 299–337; and David Mindell, Jerome
Segal, and Slava Gerovitch, “From Communications Engineering to Communica-
tions Science: Cybernetics and Information Theory in the United States, France, and
the Soviet Union,” in Science and Ideology: A Comparative History, ed. Mark Walker,
66–96 (New York: Routledge, 2003).Work on cybernetics in France includes, among
others, Celine Lafontaine, “The Cybernetic Matrix of ‘French Theory,’” Theory, Cul-
ture and Society 24 (2007): 27–46; Lydia Liu, “The Cybernetic Unconscious: Rethink-
ing Lacan, Poe, and French Theory,” Critical Inquiry 36 (2010): 288–320; Bernard
Geoghegan, “From Information Theory to French Theory: Jakobson, Lévi-Strauss,
and the Cybernetic Apparatus,” Critical Inquiry 38 (2011): 96–126. On cybernetics in
Britain, see Andrew Pickering, The Cybernetic Brain: Sketches of Another Future (Chi-
cago: University of Chicago Press, 2010).On cybernetics in East Germany, see Jérôme
Segal, “L’introduction de la cybernétique en R.D.A. rencontres avec l’idéologie marx-
iste,” Science, Technology and Political Change: Proceedings of the Twentieth International
Congress of History of Science (Liège, July 20–26, 1997) (Brepols: Turnhout, 1999), 1:
67–80.And on cybernetics in China, see Susan Greenhalgh, “Missile Science, Popula-
tion Science: The Origins of China’s One-Child Policy,” China Quarterly 182 (2005):
253–276. On cybernetics in Chile, see Medina, Cybernetic Revolutionaries.

21. I owe the term knowledge base to conversations with Richard John in 2010. See,
in particular, his related work on the political decisions that have shaped U.S. com-
munication history, Network Nation: Inventing American Telecommunications (Cam-
bridge: Harvard University Press, 2010).

22. Stephen Jay Gould, Life’s Grandeur (London: Vintage, 1997), 7.

23. Under the name “actor-network theory,” Bruno Latour has attempted to theo-
rize the concept of network as a way of retooling the historian’s method of following
the linkages across all forms of actors. See Bruno Latour’s Science in Action: How to
Follow Scientists and Engineers through Society (Milton Keynes: Open University Press,
1987). Two decades later, he deemed “the word network so ambiguous we should
have abandoned it long ago,” in Bruno Latour, Reassembling the Social: An Introduc-
tion to Actor-Network-Theory (Oxford: Oxford University Press, 2005), 129–130.

9800.indb 225 6/2/16 3:05 PM

226  Notes to Inroduction and Chapter 1

24. For more on the historical designator new media, see Benjamin Peters, “And Lead
Us Not into Thinking the New Is New: A Bibliographic Case for New Media History,”
New Media and Society 11 (1–2) (2009): 13–30.

25. Aleksandr Ya. Khinchin, “Teoria prosteishego potoka” (Mathematical Methods

of the Theory of Mass Service; more literally, Simple Stream Theory), Trudy
Matematicheskogo Instituta Steklov. 49 (1955): 3–122.

26. János Kornai, The Socialist System: The Political Economy of Communism (Prince-
ton: Princeton University Press, 1992); David Graeber, Debt: The First Five Thousand
Years (New York: Melville House, 2011), 94.

27. The field of institutional economics offers pragmatic approaches to observed

irrationalities in individual and group actions. A few standard references in the lit-
erature include Thorsten Veblen’s heterodox position in “Why Is Economics Not an
Evolutionary Science?,” Quarterly Journal of Economics 12 (1898): 373–393; Thomas
C. Schelling, Micromotives and Macrobehavior (New York: Norton, 1978); Douglass C.
North, Institutions, Institutional Change and Economic Performance (New York: Cam-
bridge University Press, 1998); Ronald Coase, “The New Institutional Economics,”
American Economic Review 88 (2) (1998): 72–74; and William Kapp, The Foundations
of Institutional Economics (New York: Routledge, 2011). For comparison to the quirki-
ness of individual decisions, see popular introductions to cognitive psychology and
behavioral psychology and economics, such as Daniel Kahnemann, Thinking Fast
and Slow (New York: Farrar, Straus, and Giroux, 2011), and Dan Ariely, Predictably
Irrational: The Hidden Forces That Shape Our Decisions (New York: HarperCollins,
2008). Compare these to recent works on the informal and violent character of post-
Soviet economics, including Alena V. Ledeneva, Russia’s Economy of Favors: Blat,
Networking and Information Exchange (New York: Cambridge University Press, 1998),
and Vadim Volkov, Violent Entrepreneurs: The Use of Force in the Making of Russian
Capitalism (Ithaca: Cornell University Press, 2002).

28. The English-language literature on tech entrepreneurs is long and popular,

including Walter Isaacson, The Innovators (New York: Simon & Schuster, 2014), and
Peter Thiel, Zero to One: Notes on Startups, or How to Build the Future (New York:
Crown Business, 2014), but very little of it to my knowledge looks beyond the West
(in particular, the west coast of the United States and the eastern Asian rim), such as
Eden Medina, ed., Beyond Imported Magic: Essays on Science, Technology, and Society in
Latin America (Cambridge: MIT Press, 2014).

Chapter 1: A Global History of Cybernetics

1. See note 20 on cybernetic literature in the introduction to this book.

2. See Wiener, Cybernetics; Bowker, “How to Be Universal: Some Cybernetic Strate-

gies, 1943–70”; Galison, “The Ontology of the Enemy”; and J. R. Pierce, “The Early

9800.indb 226 6/2/16 3:05 PM

Notes to Chapter 1  227

Days of Information Theory,” IEEE Transactions on Information Theory 19 (1) (1973):

3–8; and especially Ronald R. Kline, The Cybernetics Moment, Or Why We Call Our Age
the Information Age (Baltimore, MD: Johns Hopkins University Press, 2015).

3. Ronald R. Kline, “Where Are the Cyborgs in Cybernetics?,” Social Studies of Science
39 (3) (2009): 331–362.

4. Wiener, Cybernetics. On the curious father-son circularities between Leo’s Slavic

studies and Norbert’s cold war cybernetics, see Benjamin Peters, “Toward a Geneal-
ogy of a Cold War Communication Science: The Strange Loops of Leo and Norbert
Wiener,” Russian Journal of Communication 5 (1) (2013): 31–43.

5. This section draws on my previously published work on cybernetics, including

Bernard Geoghegan and Benjamin Peters, “Cybernetics” in The John Hopkins Guide to
Digital Media, ed. Marie-Laure Ryan et. al. (Baltimore: John Hopkins University Press,
2014), 109–112.

6. Wiener’s classic works include his technical masterpiece Cybernetics, the popular
The Human Use of Human Beings: Cybernetics and Society (Boston: Houghton Mifflin,
1950), and his deathbed lectures God and Golem, Inc.: A Comment on Certain Points
Where Cybernetics Impinges on Religion (Cambridge: MIT Press, 1964).

7. Wiener, Cybernetics, 1–25, 155–168.

8. Ibid., 16.

9. Dupuy, Mechanization of the Mind. See also John von Neumann, The Computer and
the Brain, 2nd ed. (New Haven: Yale University Press, [1958] 2000).

10. Quoted in Claus Pias, “Analog, Digital, and the Cybernetic Illusion,” Kybernetes
34 (3–4) (2005): 544.

11. Claus Pias, ed., Cybernetics-Kybernetik 2: The Macy-Conferences 1946–1953 (Berlin:

Diaphanes, 2004).

12. Steve J. Heims, The Cybernetics Group (Cambridge: MIT Press, 1991).

13. Ibid., 52–53, 207.

14. William Aspray, John von Neumann and the Origins of Modern Computing (Cam-
bridge: MIT Press, 1990).

15. David Lipset, Gregory Bateson: The Legacy of a Scientist (New York: Prentice Hall,
1980). See also Fred Turner, From Counterculture to Cyberculture (Chicago: University
of Chicago Press, 2006), 121–125.

16. Jefferson Pooley, “An Accident of Memory: Edward Shils, Paul Lazarsfeld and the
History of American Mass Communication Research,” Ph.D. diss., Columbia Univer-
sity, New York, 2006.

9800.indb 227 6/2/16 3:05 PM

228  Notes to Chapter 1

17. For more on “trading zones,” see Peter Galison, Image and Logic: A Material Cul-
ture of Microphysics (Chicago: University of Chicago Press, 1997), 44–47, 781–784,
806–807, 816–817.

18. Bill Aspray, “The Scientific Conceptualization of Information,” Annals of the His-
tory of Computing 7 (2) (1985): 117–140.

19. Claude E. Shannon, “A Mathematical Theory of Communication,” Bell Systems

Technical Journal 27 (1948): 379–423, 623–656.

20. Mirowski, Machine Dreams.

21. Paul Erickson, Judy L. Klein, Lorraine Dastone, Rebecca Lemov, Thomas Sturm,
and Michael D. Gordin, How Reason Almost Lost Its Mind: The Strange Career of Cold
War Rationality (Chicago: University of Chicago Press, 2013).

22. Claude E. Shannon, “The Bandwagon,” IRE Transactions on Information Theory 2

(1) (1956): 3. See also Pierce, “The Early Days of Information Theory”; Norbert
Wiener, “What Is Information Theory?,” IRE Transactions on Information Theory 48
(1956): 48; Ronald R. Kline, “What Is Information Theory a Theory Of? Boundary
Work among Scientists in the United States and Britain during the Cold War,” in
The History and Heritage of Scientific and Technical Information Systems: Proceedings of
the 2002 Conference, Chemical Heritage Foundation, ed. W. Boyd Rayward and Mary
Ellen Bowden, 15–28 (Medford, NJ: Information Today, 2004).

23. Arturo Rosenblueth, Norbert Wiener, and Julian Bigelow, “Behavior, Purpose,
and Teleology,” Philosophy of Science 10 (1943): 18–24.

24. Daniel Kahneman and Amos Tversky, “Prospect Theory: An Analysis of Deci-
sions under Risk,” Econometrica 47 (2) (1979): 263–291. See also Daniel Kahneman
and Amos Tversky, eds., Choices, Values and Frames (New York: Cambridge Univer-
sity Press and Russell Sage Foundation, 2000).

25. David Stark, The Sense of Dissonance: Accounts of Worth in Economic Life (Prince-
ton: Princeton University Press, 2009), 1–34.

26. The intellectual history of thought on hierarchy and its critics would fill many
shelves. That history might combine thinking on technical subordination in math-
ematics (cardinal numbers, graphs, networks, sets, type theory, programming) and
other classificatory systems; individual autonomy (Plato, Locke and Kant, Isaiah
Berlin and Charles Taylor) and sociobiological evolution; legal, ethical, and religious
thought; and pragmatism and feminism. For a helpful update on modern network
discourse, see Daniel Kreiss, Megan Finn, and Fred Turner, “The Limits of Peer Pro-
duction: Some Reminders from Max Weber for the Network Society,” New Media and
Society 13 (2) (2011): 243–259.

27. Warren S. McCulloch, “A Heterarchy of Values Determined by the Topology of

Nervous Nets,” Bulletin of Mathematical Biophysics 7 (1945): 89–93.

9800.indb 228 6/2/16 3:05 PM

Notes to Chapter 1  229

28. Ibid., 91.

29. George Dyson, Turing’s Cathedral: The Origins of the Digital Universe (New York:
Pantheon Books, 2012), 196–197, see also 7–10, 56–63.

30. John von Neumann, “Can We Survive Technology?,” Fortune (June 1955): 106–
108, 151–152.

31. For a lively discussion, see Dupuy, The Mechanization of Mind.

32. For a few examples of the French scholarly and popular presses on cybernetics
between 1946 and 1952, see Jacque Bergier, “Un plan général d’automatisation des
industries,” Les Lettres françaises (April 15, 1948): 7–8; Léon Brillouin, “Les machines
américaines,” Annales des Télécommunications 2 (1947): 331–346; Léon Brillouin,
“Les grandes machines mathématiques américaines,” Atomes 2 (21) (1947): 400–404;
Louis de Broglie, La cybernétique: théorie du signal et de l’information (Paris: Edition de
la Revue d’Optique Théorique et Instrumentale, 1951); Dominique Dubarle, “Une
nouvelle science: la cybernétique—vers la machine à gouverner?,” Le Monde, Decem-
ber 28, 1948, in P. Breton, A l’image de l’homme (Paris: Seuil, 1995), 137–138; Domi-
nique Dubarle, “Idées scientifiques actuelles et domination des faits humains,” Esprit
9 (18) (1950): 296–317. See also Jérôme Segal, Le zéro et le un: histoire de la notion sci-
entifique d’information (Paris: Syllepse, 2003).

33. Mindell, Segal, and Gerovitch, “From Communications Engineering to Commu-

nications Science.”

34. Ibid. See also Geoghegan, “From Information Theory to French Theory”; Céline
LaFontaine, “The Cybernetic Matrix of French Theory”; and LaFontaine, L’empire
cybernétique: des machines à penser à la pensée machine (Paris: Seuil, 2004).

35. Phil Husbands and Owen Holland, “The Ratio Club: A Hub of British Cyberneti-
cists,” in The Mechanical Mind in History, ed. P. Husbands, O. Holland, and M.
Wheeler, 91–148 (Cambridge: MIT Press, 2008).

36. Pickering, The Cybernetic Brain.

37. Stafford Beer, Brain of the Firm (London: Allen Lane, Penguin Press, 1972).

38. Humberto Maturana and Francisco Valera, Autopoiesis and Cognition: The Realiza-
tion of the Living (Boston: Reidel, 1980); Francisco Valera, The Tree of Knowledge: The
Biological Roots of Human Understanding (Boston: Shambhala Press, 1987); Francisco
Valera with Evan Thompson and Eleanor Rosch, The Embodied Mind: Cognitive Sci-
ence and Human Experience (Cambridge: MIT Press, 1991).

39. For more on Aleksandr Bogdanov, see his Tektologia: Vsyeobshcheiye Organizatsi-
onnaya Nauka (Tectology: Universal Organizational Science) (Moscow: Akademia Nauk,
1913–1922). See also Nikolai Krementsov, A Martian Stranded on Earth: Alexander
Bogdanov, Blood Transfusions, and Proletarian Science (Chicago: University of Chicago
Press, 2011), and J. Biggart, P. Dudley, and F. King, eds., Alexander Bogdanov and the

9800.indb 229 6/2/16 3:05 PM

230  Notes to Chapter 1

Origins of Systems Thinking in Russia (Brookfield, VT: Ashgate, 1998), and McKenzie
Wark, Molecular Red: A Theory for the Anthropocene (New York: Verso, 2015).

40. On Stefan Odobleja, see Mihai Draganescu, Odobleja: Between Ampère and Wiener
(Bucharest: Academia Republicii Socialiste Romania, 1981); Nicolae Jurcau, “Two
Specialists in Cybernetics: Stefan Odobleja and Norbert Wiener, Common and Dif-
ferent Features,” Twentieth World Congress of Philosophy (1998), accessed October 11,
2011, http://www.bu.edu/wcp/Papers/Comp/CompJurc.htm.

41. Peters, “Toward a Genealogy of a Cold War Communication Science.”

42. Michael O’Shea. The Brain: A Very Short Introduction (New York: Oxford Univer-
sity Press, 2005), 1.

43. For canonic works on Soviet science written during or soon after the cold war,
see Zhores Medvedev, Soviet Science (New York: Norton, 1978); Alexander Vucinich,
Empire of Knowledge: The Academy of Sciences of the USSR (1917–1970) (Berkeley: Uni-
versity of California Press, 1984); Graham, Science in Russia and the Soviet Union;
David Joravsky, Soviet Marxism and Natural Science, 1917–1932 (New York: Columbia
University Press, 1971). For more current materials, see Nikolai Krementsov, Stalinist
Science (Princeton: Princeton University Press, 1996); Paul R. Josephson, Totalitarian
Science and Technology (Atlantic Highlands, NJ: Humanities Press, 1996); Paul R.
Josephson, Red Atom: Russia’s Nuclear Power Program from Stalin to Today (Pittsburg:
University of Pittsburg Press, 2005); and Ethan Pollock, Stalin and the Soviet Science
Wars (Princeton: Princeton University Press, 2006).

44. Nils Roll-Hansen, The Lysenko Effect: The Politics of Science (Amherst, NY: Human-
ity Books, 2005).

45. For a thorough discussion of the politics of the label “Lysenkoism,” see William
deJong-Lambert and Nikolai Krementsov, “On Labels and Issues: The Lysenko Con-
troversy and the Cold War,” Journal of the History of Biology 45 (3) (2012): 373–388,
and especially Audra J. Wolfe, “The Cold War Context of the Golden Jubilee, or,
Why We Think of Mendel as the Father of Genetics,” Journal of the History of Biology
45 (3) (2012): 389–414. Earlier materials include David Joravsky, The Lysenko Affair
(Chicago: University of Chicago Press, 1970), and Valery N. Soyfer, Lysenko and the
Tragedy of Soviet Science (New Brunswick, NJ: Rutgers University Press, 1994).

46. Gerovitch, From Newspeak to Cyberspeak, 547–548.

47. Ibid., 120.

48. Ibid., 126.

49. Mikhail G. Iaroshevskii, “Semanticheskii idealizm: filosofiia imperialisticheskoi

reaktsii,” in Protiv filosofiia oruzhenostsev amerikano-angliiskogo imperializma, ed. T.
Oizerman and P. Trofimov (Moscow: Nauka, 1951), 100, quoted in Gerovitch, From
Newspeak to Cyberspeak, 119–121.

9800.indb 230 6/2/16 3:05 PM

Notes to Chapter 1  231

50. Gerovitch, From Newspeak to Cyberspeak, 119–121.

51. Published under the pseudonym “Materialist,” Voprosy Filisofii 5 (1953): 210–

52. Gerovitch, From Newspeak to Cyberspeak, 124–126. See also Ocherki istorii informa-
tiki v Russii, ed. D. Pospelov and Ya. Fet (Novosibirsk: Nauchnyi Tsentr Publikatsii
RAS, 1998).

53. Mark M. Rosenthal and Pavel F. Iudin, eds., Kratkiĭ filosofskiĭ slovar’, 4th ed.
(Moscow: Gospolitizdat, 1954), 236–237; also quoted in Masani, Norbert Wiener, 261.

54. Gerovitch, From Newspeak to Cyberspeak, 119.

55. Ilia B. Novik, “Normal’naia Lzhnauka” [“A Normal Pseudoscience”], Voprosy isto-
rii estestvoznaniia i tekhniki (Questions of History of Natural Science and Technology) 4
(4) (1990), quoted in Gerovitch, From Newspeak to Cyberspeak, 103.

56. For more reading on Soviet science, see note 43. On Vygotsky in particular, see
Alex Kozulin, Vygotsky’s Psychology: A Biography of Ideas (Cambridge: Harvard Uni-
versity Press, 1990).

57. Sergei N. Khrushchev, Nikita Khrushchev and the Creation of a Superpower (Univer-
sity Park: Pennsylvania State University Press, 2000), 94–96, 163–173.

58. George Paloczi-Horvath, Khrushchev: The Making of a Dictator (New York: Little,
Brown, 1960), 202.

59. Paloczi-Horvath, Khrushchev: The Making of a Dictator, 202.

60. Ibid., 80.

61. David Holloway, “Physics, the State, and Civil Society in the Soviet Union,” His-
torical Studies in Physical and Biological Sciences 100 (1) (1999): 173–192; Loren R.
Graham, “How Robust Is Science under Stress?,” What Have We Learned about Science
and Technology from the Russian Experience? (Stanford: Stanford University Press,
1998), 52–73.

62. Ivan Pavlov, Conditioned Reflexes: An Investigation of the Physiological Activity of the
Cerebral Cortex, trans. and ed. G. V. Anrep (London: Oxford University Press, 1927).

63. Josephson, New Atlantis Revisited.

64. Stanislav Boguslavski, Henryk Grenievski, and Jerzy Szapiro, “Dialogi o cyberne-
tyce,” Mysl filozoficzna 4 (14) (1954): 158–212, cited in Günther, “Cybernetics and
Dialectical Materialism of Marx and Lenin.”

65. Anatoly Kitov, “Chelovek, kotoryi vynes kibernetiku iz sekretnoi biblioteki”

[“The Man Who Brought Cybernetics Out of a Secret Library”], interview,
Komp’iuterra 43 (November 18, 1996): 44–45.

9800.indb 231 6/2/16 3:05 PM

232  Notes to Chapter 1

66. Gerovitch, From Newspeak to Cyberspeak, 173–175.

67. Ibid., 176–180.

68. Kitov, “Chelovek, kotoryi vynes kibernetiku iz sekretnoi biblioteki,” 44–45.

69. Gerovitch, From Newspeak to Cyberspeak, 183.

70. Ibid., 173. See also Sergei L. Sobolev, Anatolii I. Kitov, and Aleksei A. Lyapunov,
“Osnovnye cherty kibernetiki” [“Basic Features of Cybernetics”], Voprosy filsofii 4
(1955): 136.

71. Edwards, The Closed World, 175–208, 275–302.

72. Gerovitch, From Newspeak to Cyberspeak, 178.

73. Wiener, Cybernetics.

74. I discuss the renewability of new media in Peters, “And Lead Us Not into Think-
ing the New Is New,” and Benjamin Peters and Deborah Lubken, “New Media in
Crises: Discursive Instability and Emergency Communication,” in The Long History
of New Media, ed. David W. Park et al., 193–209 (New York: Peter Lang, 2011).

75. My thanks to Andriy Ishchenko and an anonymous reviewer for this distinction.

76. Sobolev, Kitov, and Lyapunov, “Osnovnye cherty kibernetiki,” 141.

77. Ibid., 141–146.

78. Ibid.

79. Ibid., 147.

80. See Karel Chapek’s play Rossum’s Universal Robots, trans. David Willie (Fairford:
Echo Library, 2010).

81. Sobolev et al., “Osnovnye cherty kibernetiki,” 148.

82. Ibid., 147.

83. Ibid.

84. Ibid.

85. Ibid.

86. For more on Kolman, see Loren Graham and Jean-Michael Kantor, Naming Infin-
ity: A True Story of Religious Mysticism and Mathematical Creativity (Cambridge:
Belknap Press of Harvard University Press, 2009).

87. Graham and Kantor, Naming Infinity. This fascinating account describes how
founding (transfinite) set theorists and religious mystics such as Dmitri Egorov,
Pavel Florensky, and Nikolai Luzhin in 1920s Moscow came together around the
realization that neither infinity nor God could be defined but both could be named.

9800.indb 232 6/2/16 3:05 PM

Notes to Chapter 1  233

88. For more on Lysenko, see deJong-Lambert and Krementsov, “On Labels and
Issues.” Wolfe, “The Cold War Context of the Golden Jubilee,” rethinks the accepted
positions against Lysenko laid out in Joravsky, The Lysenko Affair, and Soyfer, Lysenko
and the Tragedy of Soviet Science.

89. Graham and Kantor, Naming Infinity, 129.

90. Arnosht (Ernest) Kol’man, My ne dolzhny byli tak zhit’ [We Should Not Have Lived
That Way] (New York: Chalidze, 1982), 7, quoted in Graham and Kantor, Naming
Infinity, 130.

91. Ernest Kolman,“Shto takoe kibernetika?” [What Is Cybernetics?”], Voprosi Filos-

ophii (Akademia Nauk CCCR Institut Filosophii, Moscow) 4 (1955): 148–149.

92. Ibid., 149.

93. Wiener briefly studied at Columbia under John Dewey in 1915 and worked as a
consultant for a National Defense Research Committee–supported Statistical
Research Group based there in 1940.

94. Kolman, “Shto takoe kibernetika?,” 150–157.

95. Helmut Dahm, “Zur Konzeption der Kybernetik im dialektischen Materialis-

mus,” unpublished manuscript, 25, quoted in Günther, “Cybernetics and Dialectical
Materialism of Marx and Lenin,” 317–332.

96. David Holloway, for example, writes that “the hostile image of capitalist society,
which had played an important part in the early attacks on cybernetics, was now
turned to its defense,” in “Innovation in Science: The Case of Cybernetics in the
Soviet Union,” Science Studies 4 (1974): 316.

97. Gerovitch, From Newspeak to Cyberspeak, 180.

98. Ibid.

99. Sobolev, Kitov, and Lyapunov, “Osnovnye cherty kibernetiki,” quoted in Gero-
vitch, From Newspeak to Cyberspeak, 180.

100. Erickson et al., How Reason Almost Lost Its Mind, 272.

101. Galison, “The Ontology of the Enemy,” 228–266. Peter Galison says that “the
enemy as human-machine black box becomes us as human-machine black box.” In
Sina Najafi and Peter Galison, “The Ontology of the Enemy: An Interview with Peter
Galison,” Cabinet 12 (2003), accessed April 10, 2015, http://cabinetmagazine.org/

102. John A. Armstrong, “Sources of Administrative Behavior: Some Soviet and

Western European Comparisons,” American Political Science Review 59 (3) (1965):

103. Gerovitch, “The Cybernetics Scare and the Origins of the Internet,” 32–38.

9800.indb 233 6/2/16 3:05 PM

234  Notes to Chapter 1

104. D. G. Malcolm, “Review of Cybernetics at [sic] Service of Communism, vol. 1,”

Operations Research 11 (1963): 1012.

105. Conway and Siegelman, Dark Hero of the Information Age, 316.

106. CIA Intelligence Memorandum, No. 0757/64, “The Meaning of Cybernetics in

the USSR,” February 26, 1964, 2, also partially quoted in Flo Conway and Jim Siegel-
man, Dark Hero of the Information Age (New York: Basic, 316).

107. CIA Intelligence Memorandum, No. 0757/64, “The Meaning of Cybernetics in

the USSR,” 3.

108. Ibid., 3.

109. Ibid., 3. See also Gerovitch, “The Cybernetics Scare and the Origins of the Inter-
net,” 35.

110. Gerovitch, “The Cybernetics Scare and the Origins of the Internet.”

111. Gerovitch, “InterNyet,” 340.

112. Gerovitch, From Newspeak to Cyberspeak, 249–251.

113. Yu. Kapitonova and A. A. Letichevsky, Paradigmi i idei akademika V. M. Glush-

kova (Kiev: Naukova Dumka, 2003), 296.

114. Igor’ A. Poletaev, “O matematicheskom modelirovnanii,” Problemy kibernetiki

27 (1973): 147.

115. Igor’ A. Poletaev, “K opredeleniiu poniatiia ‘informatsiia,’” in Issledovanniia po

kibernetike, ed. A. Lyapunov (Moscow: Sovetskoe radio 1970), 212.

116. Gerovitch, From Newspeak to Cyberspeak, 216.

117. Simon Kassel, Soviet Cybernetics Research: A Preliminary Study of Organizations and
Personalities (Santa Monica: RAND, 1971), v.

118. See the titles of Wiener’s Cybernetics, or Control and Communication in the Animal
and the Machine and his The Human Use of Human Beings: Cybernetics and Society
(Cambridge: MIT Press, 1950).

119. Gerovitch, From Newspeak to Cyberspeak, 208.

120. Ibid., 209–210.

121. Ibid., 210.

122. Pospelov and Fet, Ocherki istorii informatiki v Rossii.

123. Conway and Siegelman, Dark Hero, 316.

124. Norbert Wiener, “Obschestvo i nauka,” Voprosi Filosofiii 7 (1961): 49–52.

9800.indb 234 6/2/16 3:05 PM

Notes to Chapter 1  235

125. Dirk Jan Struik, “Norbert Wiener: Colleague and Friend,” American Dialog 3 (1)
(1966): 34–37.

126. Bonnie Honig, Democracy and the Foreigner (Princeton: Princeton University
Press, 2003).

127. On the one hundred twentieth anniversary of his birth and the fiftieth anniver-
sary of his death, the IEEE held a medium-sized conference in Boston on June 24–26,
2014, titled Norbert Wiener in the Twenty-first Century, including a gathering of
biographers, former students of his, and rising scholars interested in his life and

128. Conway and Siegelman, Dark Hero, 314–316. See also Peters, “Toward a Geneal-
ogy of Cold War Communication Science.”

129. Gerovitch, From Newspeak to Cyberspeak, 154–155, 301, passim.

130. James W. Carey with John J. Quirk, “The Mythos of the Electronic Revolution,”
in Communication and Culture: Essays on Media and Society, 113–141 (New York:
Unwin Hyman, 1989).

131. For more on feedback in the Western political tradition, see Otto Mayr, Author-
ity, Liberty, and Automatic Machinery in Early Modern Europe (Baltimore: Johns Hop-
kins University Press, 1989), 144; see also Bernard Geoghegan, “The Cybernetic
Apparatus: Media, Liberalism, and the Reform of the Human Sciences,” Ph.D. diss.,
Northwestern University, Chicago, 2012.

132. Gerovitch, From Newspeak to Cyberspeak, 122.

133. Graham, Lonely Ideas, 1–4. For contrasting portraits of the local contingencies
and practices that animate laboratory work, see Galison, Image and Logic, and Latour,
Science in Action.

134. Dima Adamsky, The Culture of Military Innovation: The Impact of Cultural Factors
on the Revolution in Military Affairs in Russia, the US, and Israel (Stanford: Stanford
University Press, 2010), 37, 24–57.

135. Medina notes that “Beer was well aware of the Soviet approach to cybernetic
management, and he viewed it with open contempt.” Eden Medina, Cybernetic Revo-
lutionaries: Technology and Politics in Allende’s Chile (Cambridge: MIT Press, 2011), 63.

136. Herbert Spencer, The Principles of Sociology, 3rd ed. (New York: Westminster,
[1876] 1896), 460–462, 478–545.

137. Charles Horton Cooley, Sociological Theory and Social Research: Selected Papers of
Charles Horton Cooley (New York: Henry Holt, 1930), 6.

138. O’Shea, The Brain, 1.

9800.indb 235 6/2/16 3:05 PM

236  Notes to Chapter 2

Chapter 2: Economic Cybernetics and Its Limits

1. Regarding the mutual embeddedness of practice and theory on which this analy-
sis of Soviet economic problems rests, I take for granted (more or less following John
Dewey and other early pragmatists) that the two cannot be separated. Without prac-
tice, theory is a mere abstraction, a desiccation of thought; without theory, practice
is purposeless action. I understand theory as a form of practice, however subdued
and meditative its rootedness in modern society may be, and practice as an expres-
sion of mental purpose, an exercising of theory in a world that knows only action.
C. S. Peirce put the point thus: “Consider what effects, which might conceivably
have practical bearings, we conceive the object of our conception to have. Then, our
conception of those effects is the whole of our conception of the object.” In other
words, to consider an object is to conceive of its practical effects. To conceive, or
theorize, an object is, for the early pragmatists, also to understand the full set of its
practices and implications. With this in mind, the analysis of organizations and
economics that follows assumes that theoretical and practical judgments must be
reconcilable. For more, see John Dewey, Logic: The Theory of Inquiry, in The Essential
Dewey, vol. 2 (Bloomington: Indiana University Press, 1999), 169–179; see also
Charles Sanders Peirce, “How to Make Our Ideas Clear,” in The Essential Peirce, vol. 1
(Bloomington: Indiana University Press, 1992–1999), 132.

2. Before 1928, the Soviet Union was an indicative economy, not a command econ-
omy, meaning that the state set economic quotas but did not compel them. Richard
E. Ericson, “Command Economy,” The New Palgrave Dictionary of Economics, 2nd ed.,
ed. Steven N. Durlauf and Lawrence E. Blume (New York: Palgrave, 2008).

3. Engels wrote to Karl Kautsky in Vienna: “In any case, it will be for those people to
decide if, when and what they want to do about it, and what means to employ. I
don’t feel qualified to offer them any advice or counsel in this matter. They will
presumably be at least as clever as we are.” Friedrich Engels to Karl Kautsky in
Vienna, from Karl Marx and Friedrich Engels, Karl Marx and Frederick Engels: Selected
Correspondence (Moscow: Progress, 1975), accessed July 25, 2013, http://www.

4. Stephen F. Cohen, Bukharin and the Bolshevik Revolution: A Political Biography,

1888–1938 (New York: Oxford University Press, 1971), 93.

5. Much of the vast literature on the Soviet command economy is dated to cold war
research concerns. The part that was consulted (and sometimes critiqued) in this
work includes Mark Beissinger, Scientific Management, Socialist Discipline, and Soviet
Power (Cambridge: Harvard University Press, 1988); Peter Blau, Bureaucracy in Modern
Society (New York: Random House, 1956); Michael Ellman, Planning Problems in the
USSR: The Contributions of Mathematical Economics to Their Solution, 1960–1971 (New
York: Cambridge University Press, 1973); Michael Ellman, Socialist Planning (New
York: Cambridge, 1978); Paul R. Gregory, The Political Economy of Stalinism: Evidence

9800.indb 236 6/2/16 3:05 PM

Notes to Chapter 2  237

from the Soviet Secret Archives (New York: Cambridge University Press, 2003); Paul R.
Gregory, Restructuring the Soviet Economic Bureaucracy (New York: Cambridge Univer-
sity Press, 1990); Gregory Grossman, ed., Studies in the Second Economy of Communist
Countries: A Bibliography (Berkeley: University of California Press, 1988); Gregory
Grossman, “Notes for a Theory of the Command Economy,” Soviet Studies 15 (2)
(1963): 101–123; Gregory Grossman, “The ‘Second Economy’ of the USSR,” Problems
of Communism 26 (5) (1977): 25–40; János Kornai, The Socialist System; Alena V. Lede-
neva, Russia’s Economy of Favors: Blat, Networking, and Informal Exchange. New York:
Cambridge University Press, 1998; Alec Nove, An Economic History of the USSR,
1917–1991, 3rd ed. (New York: Penguin, 1992); Elena Osokina, Our Daily Bread:
Socialist Distribution and the Art of Survival in Stalin’s Russia, 1927–1941 (New York:
Routledge, 2003); and Alejandro Portes, Manuel Castells, and Lauren A. Benton,
eds., The Informal Economy: Studies in Advanced and Less Developed Countries (Balti-
more: Johns Hopkins University Press, 1989).

6. George M. Armstrong Jr., The Soviet Law of Property (The Hague: Martinus Nijhoff,
1983). See also John N. Hazard, Communists and Their Law: A Search for the Common
Core of the Legal Systems of the Marxian Socialist States (Chicago: University of Chi-
cago, 1969), 171–223; and Karl Marx and Friedrich Engel, “Manifesto of the Com-
munist Party” (1848) in Robert C. Tucker, ed., The Marx-Engels Reader, 2nd Edition
(New York: W. W. Norton, 1978): “the theory of the Communists may be summed
up in the single sentence: abolition of private property.”

7. David Dyker, Restructuring the Soviet Economy (New York: Routledge, 1991), 7.

8. Mark Harrison, “Soviet Economic Growth since 1928: The Alternative Statistics of
G. I. Khanin,” Europe-Asis Studies 45 (1) (1993): 141–167.

9. Estimates of the number of victims range from roughly 7 million to 14 million.

Robert Conquest, The Harvest of Sorrow (New York: Oxford University Press, 1986);
Timothy Snyder, Bloodlands: Europe between Hitler and Stalin (New York: Basic Books,
2012); Miron Dolot, Execution by Hunger: The Hidden Holocaust (New York: Norton,

10. David C. Engerman, Modernization from the Other Shore: American Intellectuals and
the Romance of Russian Development (Cambridge: Harvard University Press, 2003).

11. Anders Aslund, How Capitalism Was Built: The Transformation of Central and East-
ern Europe (New York: Cambridge University Press, 2007), 75; Noel E. Firth and James
H. Noren, Soviet Defense Spending: A History of CIA Estimates, 1950–1990 (College Sta-
tion: Texas A&M University Press).

12. Francis Spufford, in his delightful novel Red Plenty, fabricates a relatable incident
in which a brake failure sends a tractor hurtling through the wall of a crucial factory
and thereby disrupts the production of a specific large piece of machinery for
months. The disruption sends a ripple of delays and costs across the national indus-

9800.indb 237 6/2/16 3:05 PM

238  Notes to Chapter 2

tries that depend on the factory for the machine that it produces. Francis Spufford,
Red Plenty (Minneapolis, MN: Graywolf Press, 2012).

13. I. Borovitski, editorial, Pravda, October 3, 1962.

14. N. Chesnenko, “‘Obshchii iazyk’ elektronnykh mashin: Problemy kodirovaniia

dannykh,” Ekonomiceskaia gazeta (47) (1973): 10.

15. Gertrude E. Schroeder, “Organizations and Hierarchies: The Perennial Search for
Solutions,” in Reorganization and Reform in the Soviet Economy, ed. Susan J. Linz and
William Moskoff (New York: Sharpe, 1988), 6.

16. Leon Smolinski, “What Next in Soviet Planning?,” Foreign Affairs 42 (3) (1964):

17. Aleksei Kuteinikov, “Pervie proekti avtomatizatsii upravleniya sovetskoi plano-

voi ekonomikoi v kontse 1950-x I nachale 1960-x gg.—‘elektronnyi sotsializm’?,”
Ekonomicheskaya istoriya (Moscow: Trudi istoricheskogo faku’teta MGU) 15 (2011):

18. Castells, End of the Millennium, 17, see also 5–68.

19. Alex Galloway, Protocol: How Control Exists after Decentralization (Cambridge: MIT
Press, 2004), 2–28, 240–247.

20. E. G. Liberman, “Plans, Profits and Bonuses,” Pravda, September 9, 1962, quoted
in The Liberman Discussion: A New Phase in Soviet Economic Thought, ed. M. E. Sharpe
(White Plains, NY: International Arts and Science Press, 1965), 000–000.

21. John Marangos, Consistency and Viability of Socialist Economic Systems (New York:
Palgrave Macmillan, 2013), esp. chapter 5.

22. David Alexander Lax, Libermanism and the Kosygin Reform (Charlottesville: Uni-
versity of Virginia Press, 1991).

23. William Taubman, Sergei Khrushchev, and Abbott Gleason, Nikita Khrushchev
(New Haven: Yale University Press, 2000), 153–154.

24. Karl W. Ryavec, Russian Bureaucracy: Power and Pathology (New York: Rowman &
Littlefield, 2005), 227–230.

25. Gottfried Liebniz, “The Art of Discovery” (1685), in Leibniz: Selections, ed. Philip
P. Wiener (New York: Charles Scribner’s Sons, 1951), 50–58.

26. Kantorovich and von Neumann were born to middle-class Jewish families in
eastern Europe. Roy Gardner, in “L. V. Kantorovich: The Price Implications of Opti-
mal Planning,” in Socialism and the Market: Mechanism Design Theory and the Alloca-
tion of Resources, ed. Peter J. Boettke, 638–648 (New York: Routledge, 2000). Few have
satisfactorily described the forces behind the phenomenal scientific output of the
generation born between 1890 and 1930 to a tiny Jewish middle class in Hungary.

9800.indb 238 6/2/16 3:05 PM

Notes to Chapter 2  239

Members of this group include mathematician and founding computer scientist and
game theorist John von Neumann; pan-prolific mathematician Paul Erdős; Nobel
laureate and founder of holography Dennis Gabor; Nobel laureate and physicist
Eugene Wigner; early supersonic aerospace engineer Theodore von Kármán; discov-
erer of the linear accelerator, the electron microscope, and nuclear chain reaction
Leo Szilard; the primary force behind the hydrogen bomb Edward Teller; codevel-
oper of BASIC computer programming John George Kemeny; historian Oszkar Jaszi;
philosopher Georg Lukacs, economist and philosopher Karl Polanyi; author Arthur
Koesler; and composer Bela Bartok. Gabor Pallo, “The Hungarian Phenomenon in
Israeli Science,” Bulletin of the History of Chemistry 25 (1) (2000): 35–42.

27. Independent of Kantorovich, Von Neumann and George Dantzig developed

similar methods in the United States after the war.

28. To add some numbers to the basic problem: assume that a square meter of pota-
toes costs two rubles to grow and sells at six rubles and that a square meter of wheat
costs three rubles and sells at seven rubles. Given 100 square meters, the linear pro-
grammer might ask, “What proportion of potatoes and wheat will maximize reve-
nue?” In practice, programmers struggled to address massively more complicated
programs, factoring into their matrices and algorithms the constraints, costs, and
effects of dozens or hundreds of variables from pesticides, fertilizer, and soil degra-

29. Iosif V. Stalin, Voprosy leninizma, 11th ed. (Moscow: Gospolitizdat, 1951), 326.

30. Abraham S. Becker, “Input-Output and Soviet Planning: A Survey of Recent

Developments,” paper prepared for the United States Air Force Project RAND, Mem-
orandum, RM 3523-PR, March 1963, accessed July 18, 2013, http://www.dtic.mil/

31. V. S. Nemchinov, O dalneishem sovershenstvovanii planirovaniya i upravleniya

narodnym khozyaistvom. Moscow: Ekonomika, 1964: 1–74. V. S. Nemchinov, “Sotsial-
isticheskoe khozyaistvovanie i planirovanie proizvodstva,” Kommunist (1964): 5.

32. A. Birman, “Neotvratimost,” Zvezda 5 (1978): 1–5.

33. Boris Nikolaevich Malinovsky, Pioneers of Soviet Computing, ed. Anne Fitzpatrick,
trans. Emmanuel Aronie, 2010, accessed April 15, 2015, http://www.sigcis.org/files/
SIGCISMC2010_001.pdf, esp. “Personal Reminisces of Viktor Glushkov,” 34–59.

34. Adamsky, The Culture of Military Innovation, 26–31; Kapitonova and Letichevsky,
Paradigmi i idei akademika V.M. Glushkova, 164.

35. Stark, The Sense of Dissonance, 1–34, see also 35–51, 54–80.

36. Spufford, Red Plenty, 208–209.

37. Kornai, The Socialist System, 121.

9800.indb 239 6/2/16 3:05 PM

240  Notes to Chapter 2

38. Ibid., 121.

39. Ibid., 122.

40. Ibid., 122–123.

41. Castells, End of the Millennium, 24.

42. For a basic review of tolkachy and other informal mechanisms in the economy,
see Mark Beissinger, Scientific Management, Socialist Discipline, and Soviet Power (Cam-
bridge: Harvard University Press, 1988); Ledeneva, Russia’s Economy of Favors; and
Alena V. Ledeneva, Can Russia Modernize? Sistema, Power Networks and Informal Gov-
ernance (New York: Cambridge University Press, 2013).

43. Byung-Yeon Kim, “Informal Economy Activities of Soviet Households: Size and
Dynamics,” Journal of Comparative Economics 31 (3) ( 2003): 532–551.

44. Kim, “Informal Economy Activities of Soviet Households,” 532–535; Simon

Johnson, Daniel Kaufmann, and Andrei Shleifer, “The Unofficial Economy in Tran-
sition,” Brookings Papers on Economic Activity 2 (1997): 159–221.

45. Ledeneva, Russia’s Economy of Favors, 12.

46. Zbigniew K. Brzezinski, The Soviet Block: Unity and Conflict, rev. ed. (New York:
Praeger, 1960), 116, see also 115–124.

47. From Elet es Tudomany, December 24, 1952, and Rude Pravo, December 21, 1952,
quoted in Brzezinski, The Soviet Block, 114.

48. David Granick, Management of the Industrial Firm in the USSR: A Study in Soviet
Economic Planning (New York: Columbia University Press, 1955), 229.

49. Gregory, Restructuring the Soviet Economic Bureaucracy, 173. On the sticking power
of informal relations in other socially networked economies, see Mark Granovetter,
“The Strength of Weak Ties,” American Journal of Sociology 78 (6) (1973): 1360–1380.

50. Gertrude Schroeder, “The Soviet Economy on a Treadmill of Reforms,” Soviet

Economy in a Time of Change, U.S. Congress Joint Economic Committee (Washing-
ton, DC: USGPO, 1979).

51. Castells, The End of the Millennium, 24.

52. Loren R. Graham, The Ghost of the Executed Engineer: Technology and the Fall of the
Soviet Union (Cambridge: Harvard University Press, 1993), 73.

53. Thorsten Veblen, The Engineers and the Price System (New York: Huebsch, 1921).

54. Castells, The End of the Millennium, 30.

55. Erickson et al., How Reason Almost Lost Its Mind, 81–106.

9800.indb 240 6/2/16 3:05 PM

Notes to Chapter 3  241

Chapter 3: From Network to Patchwork

1. V. A. Kitov, E. N. Filinov, and L. G. Chernyak, “Anatoly Ivanovich Kitov,” accessed

May 19, 2010, http://www.computer-museum.ru/galglory/kitov.htm; and Vladimir
A. Kitov and Valery V. Shilov. “Anatoly Kitov: Pioneer of Russian Informatics,” in
History of Computing: Learning from the Past, vol. 325, ed. Arthur Tatnall (New York:
Springer, 2010), 80–88.

2. Slava Gerovitch, From Newspeak to Cyberspeak (Cambridge: MIT Press, 2002), 138–

3. Richard J. Samuels, “Rich Nation, Strong Army”: National Security and the Technologi-
cal Transformation of Japan (Ithaca: Cornell University Press, 1994), 1–32.

4. See Edwards, The Closed World, 75–115, esp. 99–100; see also Thomas Hughes’s
Rescuing Prometheus: Four Monumental Projects That Changed the Modern World (New
York: Vintage, 2000), esp. chap. 2 on SAGE and chap. 4 on ARPANET.

5. Kitov, Filinov, and Chernyak, “Anatoly Ivanovich Kitov.”

6. Gerovitch, “InterNyet,” 338–339. See also Theodore Shabad, “Khrushchev Says

Missile Can ‘Hit a Fly’ in Space,” New York Times, July 17, 1962. Marshal Rodion
Malinovsky, the minister of defense, made a similar claim more carefully several
months earlier: “the problem of destroying ballistic missiles in flight has been suc-
cessfully solved” as reported in an unnamed article in Pravda, October 25, 1961. For
imaginatively named radar networks, see Ashton B. Carter and David N. Schwartz,
Ballistic Missile Defense (Washington, DC: Brookings Institution Press, 1984), 197–

7. Recent technology commentators and scholars have enthused about the analog
update of “cognitive surplus” that can be made available over collaborative peer-
based computer networks. See, for example, Clay Shirky, Cognitive Surplus: Creativity
and Generosity in a Connected Age (New York: Penguin Books, 2010), and Yochai Ben-
kler, The Wealth of Networks: How Social Production Transforms Markets and Freedom
(New Haven: Yale University Press, 2006), 35–132.

8. Anatoly Kitov, Electronnie tsifrovie mashini [Electronic Ciphered Machines] (Moscow:

Radioeletronika Nauka, 1956).

9. Charles Eames and Ray Eames, A Computer Perspective: Background to the Computer
Age (Cambridge: Harvard University Press, 1973), 64, 96–97.

10. Marc Raeff, The Well-Ordered Police State: Social and Institutional Change through
Law in the Germanies and Russia, 1600–1800 (New Haven: Yale University Press,
1983); Jacob Soll, The Information Master: Jean-Baptiste Colbert’s Secret State Intelligence
System (Ann Arbor: University of Michigan Press, 2009).

9800.indb 241 6/2/16 3:05 PM

242  Notes to Chapter 3

11. Eduard A. Meerovich, “Obsuzhdenie doklada professor A. A. Liapunova ‘Ob

ispol’zovanii matematicheskikh mashin v logicheskikh tseliakh” (1954), in Ocherki
istorii informatiki, ed. D. Pospelov and Ya. Fet (Novosibirsk: Nauchnyi Tsentr Pub-
likatsii RAS, 1998), accessed March 20, 2010, http://ssd.sscc.ru/PaCT/history/early.

12. Isaak C. Bruk, “Elektronnie vyichislitel’nie mashinyi—na sluzhbu narodnomu

khozyaistvu,” Kommunist 7 (1957): 127.

13. A. I. Kitov, “Pis’mo zamestitelya nachal’nika VTs Minoboroni SSSR A.U. Kitova v
TsK KPSS N.S. Khrushchyovu ot 7 Anvarya 1959 goda,” Politekhnicheskii museu RF,
fond “Kitov Anatolii Ivanovich,” f. 228, edinitsa khraneniya KP27189/20.

14. Anatoly Kitov, “Rol’ akademika A. I. Berga v razvitii vyichislitel’noi tekhniki I

avtomatizirovannikh system upravleniya” (“The Role of Academician A. I. Berg in
the Development of Computational Technology and Automated Management Sys-
tems”), in Put’ v bol’shuyu nauku: Akademik Aksel’ Berg [Pathway to Big Science: Acade-
mician Aksel Berg] (Moscow: Nauka, 1988), accessed March 20, 2010, http://www

15. Anatoly Kitov, “Letter to Khrushchev,” January 7, 1959, 2, Politechnical Museum

of the Russian Federation, Collection Kitov, Anatoly Ivanovich, file 228, unit of stor-
age KP27189/20.

16. Ibid., 1–2.

17. Anatoly Kitov, “Chelovek, kotoryi vynes kibernetiku iz sekretnoi biblioteki”

[“The Man Who Brought Cybernetics out of a Secret Library”], Komp ‘iuterra 18 (43)
(1996): 44–45.

18. Ibid.

19. Aleksei Kuteinikov, “Pervie proekti avtomatizatsii upravleniya sovetskoi plano-

voi ekonomikoi v kontse 1950-x I nachale 1960-x gg.—‘elektronnyi sotsializm’?”
Ekonomicheskaya istoriya, 124–138 (Moscow: Trudi istoricheskogo faku’teta MGU 15,
2011), 109.

20. For a hagiographic biographical blurb (in Russian) on Konstantin Konostanti-

novich Rokossowski, see his memorial site, accessed July 25, 2012, http://www

21. Lewis Mumford, Technics and Civilization (New York: Harvest Book, 1934).

22. Kitov, “Chelovek, kotoryi vynes kibernetiku iz sekretnoi biblioteki.”

23. For more on Soviet military, see Roger R. Reese, The Soviet Military Experience
(New York: NP, 2000), and William E. Odom, The Collapse of the Soviet Military (New
Haven: Yale University Press, 1998).

24. Kitov, ““Chelovek, kotoryi vynes kibernetiku.”

9800.indb 242 6/2/16 3:05 PM

Notes to Chapter 3  243

25. A sample of Kitov’s publications relevant to the EASU can be found here,
accessed July 25, 2013: http://www.kitov-anatoly.ru/naucnye-trudy/perecen-

26. Berg, Kitov, and Lypunov, “O vozmozhnostyakh avotmatizatsii upravleniya

narodniym kozyaistvom.”

27. Postanovlenie TsK KPCC I Soveta Ministrov SSSR, “Ob uluchshenii rukovodstva
vnedreniem bychislitel’noi tekhniki i avtomatizirovannikh system upravleniya v
narodnoe khozyaistvo,” May 21, 1963, Gosudarstvenni archive (GA RF): f. 5446, o.
106, d. 1324, l. 160–172. This document is published in full for the first time in
Aleksei Viktorovich Kuteinikov, “Proekt Obshchegosudarstvennoi avtomatizirovan-
noi sistemi upravleniya sovetskoi ekonomikoi (OGAS) i problem ego realizatsii v
1960–1980-x gg.”

28. For example, one year earlier, Boris Pasternak was ridiculed in literary circles for
being awarded and then declining the Nobel Prize in literature. It is another case of
being punished for appealing too successfully to unapproved authorities. For more
on the campaign against Pasternak and others, see Solomon Volkov, The Magical
Chorus: A History of Russian Culture from Tolstoy to Solzhenitsyn (New York: Knopf,
2008), 195–196.

29. Boris Nikolaevich Malinovskii, Istoriaa vychislitel’noi tekhniki (Kiev: Gorobets,

2007), 197–207, accessed April 15, 2015, http://lib.ru/MEMUARY/MALINOWSKIJ.

30. J.C.R. Licklider, “Man-Machine Symbiosis,” IRE Transactions on Human Factors in

Electronics HFE-1 (March 1960): 4–11, see section 5.1, “Speed Mismatch between
Men and Computers,” accessed July 25, 2013, http://groups.csail.mit.edu/medg/

31. This phrase comes from a CIA declassified document that the author received via
a Freedom of Information Act request. For the phrase “unified information net-
work,” see CIA declassified documents by John J. Ford: A. “The Meaning of Cyber-
netics in the USSR,” Intelligence Memorandum No. 0757/64, February 26, 1964,
1–10, esp. 1. See also B. “The Cybernetic Approach to Education in the USSR,” Scien-
tific Intelligence Memorandum No. 464693, May 25, 1964; C1. “The Soviet Applica-
tions of Cybernetics in Medicine: 1. Medical Diagnosis,” Scientific and Technical
Intelligence Report No. 464692, September 15, 1966; C2. “2. Artificial Limbs,” Scien-
tific and Technical Intelligence Report No. 464691, May 10, 1967; D. “Major Devel-
opments in the SovBloc Cybernetics Programs in 1965,” Scientific and Technical
Intelligence Report No. 464694, October 3, 1966, 1–33.

32. Robert A. Divine, The Sputnik Challenge: Eisenhower’s Response to the Soviet Satellite
(New York: Oxford University Press, 1993), 104, 111–112, 146–155.

9800.indb 243 6/2/16 3:05 PM

244  Notes to Chapter 3

33. Janet Abbate, Inventing the Internet (Cambridge: MIT Press, 1999); see also Peter
H. Salus, ed. The ARPANET Sourcebook (Charlottesville, VA: Peer-to-Peer Communi-
cations LLC, 2008).

34. Another “humor-neutic” reading might have God speaking Hebrew through the
wires (lo means “no” or “not” in modern Hebrew).

35. Abbate, Inventing the Internet, 75–77.

36. Audra J. Wolfe, Competing with the Soviets: Science, Technology, and the State in
Cold War America (Baltimore: Johns Hopkins University Press, 2013), 49–50, ibid.,
esp. chaps. 2 and 3; Stuart W. Leslie, The Cold War and American Science: The Military-
Industrial-Academic Complex (New York: Columbia University Press, 1993), 203–231.

37. Kristie Mackrasis, Seduced by Secrets: Inside the Stasi’s Spy-Tech World (New York:
Cambridge University Press, 2014), 23, 133, 139, esp. 112–140.

38. Judy O’Neill, “Interview with Paul Baran,” Charles Babbage Institute, OH 182,
March 5, 1990, Menlo Park, CA, accessed April 15, 2015, http://www.gtnoise.net/

39. Ibid.; see also Stewart Brand, “Founding Father,” Wired 9 (3) (1991), accessed
April 15, 2015, http://archive.wired.com/wired/archive/9.03/baran_pr.html.

40. Brand, “Founding Father.”

41. Ibid.

42. Bradley Voytek, “Are There Really as Many Neurons in the Human Brain as Stars
in the Milky Way?,” Nature (Scitable blog, May 20, 2013), accessed April 15, 2015,

43. Katie Hafner and Matthew Lyon, Where the Wizards Stay up Late (New York:
Simon & Schuster, 1996), 64.

44. James Carey describes the process of communication as “models of and for real-
ity that make the world apprehensible” in “A Cultural Approach to Communica-
tion,” Communication as Culture: Media and Society (New York: Unwin Hyman, 1989),

45. Aleksandr Kharkevich, “Informatsia i tekhnika” [“Information and Technol-

ogy”], Kommunist 17 (1962): 94.

46. Ibid. For an example of his earlier and largely technocratic information theory
work, see Aleksandr A. Kharkevich, “Basic Features of a General Theory of Commu-
nication,” Radiotekhnika [Radio Engineering] 9 (5) (1954). For the CIA document, see
Conway and Siegelman, Dark Hero of the Information Age, n. 318.

47. Shannon, “A Mathematical Theory of Communication.”

48. Ibid., 102.

9800.indb 244 6/2/16 3:05 PM

Notes to Chapters 3 and 4  245

49. Kharkevich, “Informatsia i teckhnika,” 102.

50. Ibid., 94.

51. For the first public formulation of Moore’s law, see Gordon E. Moore, “Cram-
ming More Components onto Integrated Circuits,” Electronics 38 (8) (1965): 114–

52. For the first systematic work to treat knowledge as an economic measure and
resource and thus to anticipate the accounting of postindustrial information and
service sectors, see Fritz Machlup, The Production and Distribution of Knowledge in the
United States (Princeton: Princeton University Press, 1962).

53. Kharkevich, “Informatsia i tekhnika,” 102.

54. Ibid., 102.

55. Ibid., 103.

56. Ibid., 102.

57. N. I. Kovalev, “Doklad o rabote i perspektivakh razvitiya VTs pri Goseko-

nomsovete” [“Report about the Work and Perspectives of the Development of Infor-
mation Technology in the Gosekonomsovet”], July 23, 1962, Rossiiskii
gosudarstvenniyi arkhiv ekonomiki (RGAE) [Russian State Archive of Economics],
Moscow, f. 9480, o. 7, d. 466, l. 77–97, quoted in Kuteinikov, “Pervie proekti,” 134
n. 3.

58. Ibid., quoted in Kuteinikov, “Pervie proektyi,” 134–135.

59. For more on the cultural complications of automation as a Soviet concept, see
Slava Gerovitch, “Human-Machine Issues in the Soviet Space Program,” Critical
Issues in the History of Spaceflight, ed. Steven J. Dick and Roger D. Launius, 107–140
(Washington, DC: NASA History Division, 2006).

60. Rasmus Kleis Nielsen, “Democracy,” in Digital Keywords: A Vocabulary for Infor-
mation Society and Culture, ed. Benjamin Peters (Princeton: Princeton University
Press, under review), accessed April 15, 2015, http://culturedigitally.org/2014/05/
democracy-draft-digitalkeywords. See also John Keane, The Life and Death of Democ-
racy (New York: Simon & Schuster, 2009).

Chapter 4: Staging the OGAS, 1962 to 1969

1. John Lewis Gaddis, The Cold War: A New History (New York: Penguin Press, 2005),

2. V. Glushkov, “Kibernetika, progress, budushchee,” Literaturnaya Gazeta, Septem-

ber 25, 1962, 1–3.

9800.indb 245 6/2/16 3:05 PM

246  Notes to Chapter 4

3. “Voprosi Strukturi, Organizatsii i sozdaniya edinoi gosudarstvennoi seti

vyichislitel’nikh tsentrov EGSVTs,” Rossiiskii gosudarstvennyi arkhiv ekonomiki
(RGAE), f. 9480, o. 7, d. 1227, l. 82–102, reproduced in full in the appendix to Alek-
sei Viktorovich Kuteinikov, “Proekt Obshchegosudarsvetnnoyi avtomatizirovannoi
sistemi upravleniya sovetskoi ekonomikoi (OGAS) i problem ego realizatsii v 1960–
1980-x gg,” Ph.D. diss., Moscow State University, 2011.

4. Aleksei Viktorovich Kuteinikov, “Proekt avtomatizirovannoi sistemi upravleniya

sovetskoi ekonomikoi (OGAS) i problem ego realizatsii v 1960–1980” [“The project
of all-state automated system of management of Soviet economics (OGAS) and the
problem of its realization in 1960–1980”], Ph.D. diss., Moscow State University,

5. Vincent Mosco, To the Cloud: Big Data in a Turbulent World (New York: Paradigm
Publishers, 2014).

6. Gerovitch, From Newspeak to Cyberspeak, 283.

7. Gerovitch, “InterNyet,” 341; Medina, Cybernetic Revolutionaries, 75.

8. Viktor Glushkov, uncollected archives in the closet of the main office, Institute of
Cybernetics, room 804, Kiev, Ukraine.

9. Kuteinikov, “Proekt Obshchegosudarstrnnoi.”

10. Author interview with Vera Glushkov, May 14, 2012.

11. Malinovsky, Pioneers in Soviet Computing, 31–34.

12. Kapitonova and Letichevsky, Paradigmi i idei akademika V. M. Glushkova, 225–


13. Ibid.

14. Ibid., 142–144.

15. Malinovsky, Istoriya vyicheslitel’noi tekhniki, 92–93.

16. Kapitonova and Letichevski, Paradigmi i idei akademika V. M. Gluschkova, 316–

317, see also 296–317.

17. Ibid.

18. From Gluskov’s unpublished memoirs “Vopreki Avtoritetam” [“Despite the

Authorities”], accessed April 15, 2015, http://lib.ru/MEMUARY/MALINOWSKIJ/5.

19. Unsourced quote on the title screen accessed April 15, 2015, http://ogas.kiev.ua.

20. Aleksandr Ivanovich Stavchikov, “Romantika pervyikh issledovanii i proektov i

ikh protivorechnaya sud’ba” [“Romanticism of Early Research and Projects and
Their Contradictory Fate”], unnamed, unpublished history of Central Economic

9800.indb 246 6/2/16 3:05 PM

Notes to Chapter 4  247

Mathematical Institute (CEMI), Moscow, read in person and returned May 2008,
chap. 2, 17. See CEMI-RAS Archive in bibliography.

21. Anatoly Kitov, “Kibernetika i upravlenie narodnym khoziastvom” [“Cybernetics

and the Management of the National Economy”], Kibernetiku—na sluzhbu Kommu-
nism (Cybernetics: In the Service of Communism), ed. Aksel’ Berg, 1 (1961): 207, 216.

22. Eden Medina, Cybernetic Revolutionaries: Technology and Politics in Allende’s Chile
(Cambridge: MIT Press, 2011), 35, 34–39, 75–76.

23. Malinovsky, Pioneers of Soviet Computing, including excerpts of Glushkov’s

unpublished memoirs, “Vopreki avtoritetam” [“Despite the Authorities”], accessed
April 15, 2015, http://lib.ru/MEMUARY/MALINOWSKIJ/5.htm.

24. An early mention in English of the “paperless office” can be found in “The Office
of the Future,” Business Week 30 (2387) (1975): 48–70. See also Abigail Sellen and
Richard Harper, The Myth of the Paperless Office (Cambridge: MIT Press, 2003); Paul A.
Marolla et al., “A Million Spiking-Neuron Integrated Circuit with a Scalable Com-
munication Network and Interface,” Science 8 (345) (2014): 668–673, accessed April
15, 2015, http://www.sciencemag.org/content/345/6197/668.

25. Kapitonova and Letichevsky, Paradigm i idei akademika V. M. Glushkova, 18.

26. Ibid., 18.

27. Their friendship eventually became a family relationship. In the 1980s, Kitov’s
son Vladimir married Glushkov’s oldest daughter, Olga, who raised a grandson
named Viktor. Glushkov’s youngest daughter, Vera, also named Glushkov’s grand-
daughter Viktoria.

28. Author’s interview with Boris Nikolaevich Malinovsky, April 7, 2012.

29. The title of Glushkov’s last scholarly book was Fundamentals of Paperless Infor-
matics. Viktor Glushkov, Osnovi bezbumazhnoi informatiki (Moscow: Nauka, 1982),

30. Quoted in Gerovitch, “InterNyet,” 345; Viktor Glushkov, Kibernetika,

vychislitel’naia tekhnika, informatika. Izbrannye trudy (Kiev: Naukova dumka, 1990),

31. Gerovitch, “InterNyet,” 342–345.

32. Vladislav Zubok, Zhivago’s Children: The Last Russian Intelligentsia (Cambridge:
Harvard University Press, 2009), 275.

33. At least two politically distinct scholars have made this same basic point force-
fully in the last decade: Niall Ferguson, The Ascent of Money: A Financial History of the
World (New York: Penguin Group, 2008); Graeber, Debt: The First Five Thousand

9800.indb 247 6/2/16 3:05 PM

248  Notes to Chapter 4

34. Glushkov, “Shto skazhet istoria,” 3, accessed April 15, 2015, http://ogas.kiev.ua/

35. Author’s interview with Vera Viktorevna Glushkov, April 30, 2012.

36. Gerovitch, “InterNyet.”

37. Boris Nikolaevich Malinovsky, Vechno Khranit [Store Eternally] (Kiev: Gorobets,
2007), 58.

38. These details are summarized from four documents in the archival materials in
Viktor M. Glushkov’s personal files, box 18, folder 1, documents 12, 119, 122, and
123 inclusive, at the Archive and Special Collections, National Academy of Sciences
of Ukraine, Kiev. Nancy Ries also examines the culture of institutional authorities as
a form of moral power in Russian Talk: Culture and Conversation during Perestroika
(Ithaca: Cornell University Press, 1997), 88–89.

39. See the memoir of the leading participants in Lebedev’s team: Sergei Lebedev,
Lev Dashevsky, and Ekaterina Shkabara, “Malaya elektronnaya shchyotnyaya mash-
ina” [“Small Electronic Digital Computer”], 1952, accessed April 15, 2015 http://
it-history.ru/images/a/af/SALebedev_MESM.pdf; Malinovsky, Istoria vyichislitel’noi
tekhniki v litsakh, 33–34.

40. See the only known other secondary document on Cybertonia, Vera V. Glush-
kova and Sergei A. Zhabin, “Virtualnaya strana Kibertonia v Institute Kibernetiki
(60–70 gg, XX vek),” in Ukrainia i svit: gumanitarno-tekhnicheska elita ta sotsialnyi
progress: tezi dopov [Ukraine and the World: Humanitarian-Technical Elite and Social
Progress], supplementary theses, International Scientific-Theoretical Conference for
Students and Graduate Students, April 4–5, 2012 (Kharkiv: NTU Kharkiv, 2012),

41. Personal correspondence with Vera Viktorevna Glushkova, February 28, 2012.

42. Public press on Cybertonia includes clippings from “Vechirnii Kiiv,” 305 (5624)
(December 31, 1962): 3, and “Vechirniy Kiev,” 309 (6588) (December 31, 1965): 2–3.
In parody of and in the same font as Vechernii Kiev, the group also issued its own
Vechernyi Kyber as the “newspaper of the council of robots” 1 (1) (1966). See also
“Podorozh v Krainu Kibertonii” [“Travel to the Country of Cybertonia”], Kievskii
Komsomoltsyi 1 (1014) (August 1, 1963): 2–3, and A. Voloshin, “Kibertonia-65,”
Vechirnii Kiev (February 16, 1965): 2. All documents are retained in the author’s per-
sonal archives.

43. Vera Viktorevna Glushkova, “Dorogoi chitatel’, dobro pozhalovat’ v ‘kiber-

toniyu’!,” Cybertonia 1 (1) (2012): 2, accessed April 15, 2015, http://miratechgroup.

44. For references on jazz in the Cold War and the Soviet Union, see S. Frederick
Starr, Red and Hot: The Fate of Jazz in the Soviet Union, 1917–1991 (New York: Oxford

9800.indb 248 6/2/16 3:05 PM

Notes to Chapter 4  249

University Press, 1994); Lisa E. Davenport, Jazz Diplomacy: Promoting America in the
Cold War Era (Jackson: University Press of Mississippi, 2009); and Penny von Eschen,
Satchmo Blows up the World: Jazz Ambassadors Play the Cold War (Cambridge: Harvard
University Press, 2004).

45. The citation for the 1965 report is this: “Rukovoditel’ temi: incognito ispolniteli:
khoteli byi ostat’sya neizvestnyimi, khotya byi dlya knachal’stva,” Otchyot laboratorii
chitayushchikh avtomatov za 1964–1965 g.g. (Kiev: Akademiya Nayk ukrainskoi ssr
institute kibernetiki AN USSR, Laboratoriya chitayushchikh avtomatov, 1965).

46. Quote taken from unmarked document in personal archives of Vera Viktorevna
Glushkova, Kiev, Ukraine.

47. Several scholarly works have drawn critical attention to the gendered perfor-
mance in technical expertise, cyborg imagery, and counterculture, although none to
my knowledge have done so in the Soviet context. For more, see Janet Abbate,
Recording Gender: Women’s Changing Participation in Computing (Cambridge: MIT
Press, 2012); Donna Haraway, “A Cyborg Manifesto: Science, Technology, and
Socialist-Feminism in the Late Twentieth Century,” Simians, Cyborgs and Women:
The Reinvention of Nature (New York: Routledge, 1991), 149–181; Nathan Engs-
menger, The Computer Boys Take Over: Computers, Programmers, and the Politics of
Technical Expertise (Cambridge: MIT Press, 2010); Turner, From Counterculture to
Cyberculture, 76–77, 300–305.

48. In an unsent October 1961 letter, Nemchinov proposes an “institute of eco-

nomic cybernetics.” H. S. Khrushchyev, Doklad na XXII s’esde KPSS 18 Okctober
1961, RAN archives, CEMI, 1959, 1, 7, 124, 2.

49. Letter from Vasily Nemchinov to the Ministry of Finance: “V ministerstvo

finansov SSSR” [“To the Ministry of Finance, USSR”], RAN archives, CEMI, 1959, 1,
7, 125, 11.

50. Ibid.

51. Kassel, Soviet Cybernetics Research, 94–96.

52. Nemchinov letter, “V ministerstvo finansov SSSR.”

53. Ibid.

54. Yuri Gavrilets, interviewed by author, CEMI, Moscow, August 20, 2008.

55. Nemchinov letter, “V ministerstvo finansov SSSR.”

56. Nikita S. Khrushchev, “Doklad na XXII S”esde KPSS” [“Concluding Speech”],

Twenty-second Congress, October 18, 1961, accessed March 19, 2010, http://www.

9800.indb 249 6/2/16 3:05 PM

250  Notes to Chapter 4

57. Letter from V. S. Nemchinov to the Bureau of the Division of Economic, Philo-
sophical, and Legal Sciences, Academy of Sciences, USSR, December 11, 1961, RAN
Archives, 1959, 1, 7, 125, 11.

58. Report signed by V. S. Nemchinov, “Dokladnaya zapiska v Otdelenii, XXII s”ezd

KPSS” [“Division Report Notes, Twenty-second Congress of the Communist Party”],
November 17, 1961, RAN Archive, CEMI, 1959, 1, 6, 106.

59. Document signed by V. S. Nemchinov, “V Byuro otdeleniya ekonomicheskikh,

filosophskikh I provavikh nauk AN CCCP” [“To the Office of the Division of Eco-
nomic, Philosophical, and Legal Sciences, the Academy of Sciences, USSR”] Septem-
ber 17, 1960, RAN Archive, CEMI, 1960.

60. Nikolai Fedorenko, Vestnik Akademii Nauk, SSSR [Herald of the Academy of Sciences,
USSR] 10 (1964): 3–14.

61. Kassel, Soviet Cybernetics Research, 98.

62. Interview with Yuri Gavrilets by the author, CEMI, Moscow, August 20, 2008.
Data taken from the report titled “Doklad Akademika N. I. Fedoreko na yubileim
zasedankii posvyashennoim 10-leniyu of Ts.E.M.I.” [“Report by Academician N. I.
Fedorenko on the Ten-Year Anniversary of CEMI”], CEMI archives, RAN, May 1973,
1959, 1, 403, 262.

63. Interview with Yuri Gavrilets by the author, CEMI, Moscow, August 20, 2008.

64. George Simmel’s Philosophie des Geldes is a classic account of the form, not the
value, of economic objects. As Simmel notes, “we may not describe exchangeability
as a likeness of value that belongs objectively to things, but we must recognize like-
ness of value as simply a name for the exchangeability.” George Simmel, Philosophie
des Geldes [The Philosophy of Money] (Leipzig: Duncker & Humblot, 1900), 46.

65. Simmel writes, for example, that “we may not describe exchangeability as a like-
ness of value that belongs objectively to things, but we must recognize likeness of
value as simply a name for the exchangeability.” Ibid., 46.

66. Nikolai Fedorenko, Vspominaya proshloe, vzglyadivaya v budushchee [Remembering

the Past, Looking into the Future] (Moscow: Nauka, 1999), 179.

67. Ibid., 179.

68. David Stark, The Sense of Dissonance: Accounts of Worth in Economic Life (Prince-
ton: Princeton University Press, 2009), 19–31.

69. Fedorenko, Vspominaya proshloe, 209–214.

70. Informational document by Nikolai Fedorenko, “Istoricheskaya spravka o

geyatel’nosti instituta s 1963 po 1966 g. by Director Akad. Fedorenko” [“Historical
Information about the Activity of the Institute from 1963 to 1966 by Director Aca-
demician Fedorenko”], CEMI archives, RAN, 1959, 1, 101, index 170.

9800.indb 250 6/2/16 3:05 PM

Notes to Chapter 4  251

71. Kassel, Soviet Cybernetics Research, 87.

72. Homepage for Central Economic Mathematical Institute: “About CEMI” section,
accessed April 15, 2015, http://www.cemi.rssi.ru/about/how/?section=about_link.

73. Eric P. Hoffmann and Robbin F. Laird. Technocratic Socialism: The Soviet Union in
the Advanced Industrial Era. (Durham: Duke University Press, 1985), 116.

74. Ibid., 114.

75. Gerovitch, From Newspeak to Cyberspeak, 272.

76. Ibid.

77. Unsourced quote, accessed April 15, 2015, http://ogas.kiev.ua.

78. The original document proposing the EGSTVs in 1963 can be found here: Post-
anovlenie Komitet KPCC I Sovet Ministrov SSSR, “Ob uluchshenii rukovodstva vne-
dreniem vyichislitel’noi tekhniki i avtomatizirovannikh system upravleniya v
narodnoe khozyaistvo,” May 21, 1963, no. 564, Kremlin, Moscow. This document
was published for the first time in Aleksei Viktorovich Kuteinikov, “Proekt
Obschegosudarstvennoi avtomatizirovannoi sistemi upravleniya osvetskoi eko-
nomikoi (OGAS) I problem ego realizatii v 1960–1980-x gg,” PhD dissertation,
Moscow State University, Moscow, 2011. In addition, the 1967 proposal approving
the regional Ukrainian GSTVs Project in 1967 can be found in Postanovleniya soveta
ministrov Ukrainian SSR, “O vnedrenii avtomatizirovannikh system upravleniya s
premeneniem vyichislitel’noi tekhniki,” May 21, 1967, no. 338, and established by
subsequent order of the Minister of Black Metallurgy, Ukrainian SSR, on March 11,
1968, no. 68.

79. Gerovitch, “InterNyet,” 344. See also Iuliia Kapitonova and Aleksandr Let-
ichevsky, Paradigmy i idei akademika V. M. Glushkova (Kiev: Naukova Dumka, 2003),

80. Malinovskii, Istoriia vychislitel’noi tekhniki, 162.

81. Christopher Felix McDonald, “Building the Information Society: A History of

Computing as a Mass Medium,” Ph.D. diss., Princeton University, 2011.

82. Letter from the director of the CSA V. N. Starovskii to the director of the GK
KNIR K. N. Rudnev, November 2, 1963, published in the appendix of Kuteinikov,
“Proekt Obshchegosudarstvennoi.”

83. V. P. Derkach, ed. Akademik V. M. Glushkov—pioneer kiberniki (Kiev: Yunior,

2003), 324.

84. Kathryn M. Bartol, “Soviet Computer Centres: Network or Tangle?,” Soviet Stud-
ies 23 (4) (1972): 608–618.

85. Malinovsky, Pioneers, xxx–xxxii.

9800.indb 251 6/2/16 3:05 PM

252  Notes to Chapters 4 and 5

86. Ibid., 33.

87. Ibid., 165.

88. Malinovsky, Store Eternally, 61–62.

Chapter 5: The Undoing of the OGAS, 1970 to 1989

1. Gerovitch, “InterNyet,” 343.

2. Glushkov, “Shto skazhet istoria?”

3. Malinovksy, Vechno Khranit, 61.

4. Gerovitch, From Newspeak to Cyberspeak, 280.

5. K. N. Rudnev, “Vyichislitel’naya tekhnika v narodnom khozyyaistve,” Izvestiya,

September 4, 1963, cited in Kuteinikov, “Pervie proekti,” 136.

6. Malinovsky, Istoriia vychislitel’noi tekhniki v litsakh [History of Computing Technology

in Personalities], reproduces a transcription of Glushkov’s dictated memoirs, Vopreki
Avtoritetam [Despite the Authorities], accessed April 15, 2015, http://lib.ru/MEMUARY/
MALINOWSKIJ/5.htm, and in partial English translation in “Academician Glush-
kov’s ‘Life Work,’” accessed April 15, 2015, http://en.uacomputing.com/stories/

7. Glushkov, Vopreki Avtoritetam.

8. Ibid.

9. Ibid.

10. For more on the proposed structure of the OGAS, see Martin Cave, Computers
and Economic Planning: The Soviet Experience (New York: Cambridge University Press,
1980), 13–15.

11. Malinovskii, Istoriia vychislitel’noi tekhniki v litsakh, 43–44.

12. Viktor M. Glushkov, “Dlya vsei strani,” Pravda, December 13, 1981. See also
Malinovsky, Vechno Khranit, 64, cf. 65. Other bibliographies suggest strani or
“nation,” not “state,” is correct, although this remains unconfirmed.

13. Kuteinikov, “Pervie proekti,” 97.

14. Gerovitch, “InterNyet,” 345–346.

15. Kuteinikov, “Pervie proekti,” 101.

16. Ibid., 119.

9800.indb 252 6/2/16 3:05 PM

Notes to Chapter 5  253

17. The ASUification (or ASUchivaniye) of the Soviet Union amounted to, the work-
ers joked in Russian, the bitchification of the country because in Russian
ASUchivaniye shares the same root as the swearword suka.

18. Malinovksy, Istoriya vyichisletel’nikh tekhniki v litsakh, 91, also 84–93.

19. Malinvosky, Vechno Khranit, 66–67.

20. Beissinger, Scientific Management, 249.

21. Glushkov, “Vopreki Avtoritetam.”

22. Viktor Glushkov, “Ten Billion Accountants Needed,” RAND Report on Soviet
Cybernetics 2 (3) (1972): 72–73, accessed April 15, 2015, http://www.rand.org/

23. Gerovitch, From Newspeak to Cyberspeak, 280.

24. Hughes, Rescuing Prometheus.

25. Eric P. Hoffmann and Robbin F. Laird, Soviet Technocratic Socialism: The Soviet
Union in the Advanced Industrial Era (Durham: Duke University Press, 1985), 115.

26. Ibid., 115.

27. Ibid., 116.

28. Ibid., 116–117.

29. Ibid., 116.

30. Viktor Glushkov, “Zabetniye myislic dlya tekh, kto ostaetsya,” January 10, 1982,
Akademik Glushkov—pioneer kibernetiki (Kiev: n.p., 2003), accessed April 15, 2015,

31. Aleksandr Ivanovich Stavchikov, “Romantika pervyikh issledovannii i proektov i

ikh protivorechnaya sud’ba” [“Romanticism of Early Research and Projects and
Their Contradictory Fate”], in an unnamed, unpublished history of the Central Eco-
nomic Mathematical Institute (TsEMI), chap. 2, Moscow, accessed 2008, 17. (See
CEMI-RAS archive in bibliography.)

32. Stavchikov, “Romantika,” 1–2.

33. Ibid., 16–17.

34. Ibid., 17.

35. In The End of the Millennium, Manuel Castells blames the incompatibility of a
vertical statist hierarchy with horizontal information networks for the collapse of
the Soviet Union, even while identifying ways that the Soviet Union did not behave
as such a structure. Simultaneously, Gerovitch, in “InterNyet,” claims that “Soviet
cyberneticians envisioned an organic, self-regulating system, but paradoxically they

9800.indb 253 6/2/16 3:05 PM

254  Notes to Chapter 5

insisted on building it by decree from above.” Although this claim is not wrong, it
misses his earlier point that, aside from having no other option, the top-down
system did not behave as a self-regulating hierarchy. The argument offered here
looks to describe the same administrative challenges by using terms like heterarchy,
which cuts a middle way through top-down and bottom-up, horizontal and vertical
network structural discourse. Castells, The End of the Millennium, 26–37, 61–66; Gero-
vitch, “InterNyet,” 347.

36. David Edmonds and John Eidinow’s popular Bobby Fisher Goes to War: How a
Lone American Star Defeated the Soviet Chess Machine (New York: Harper Perennial,

37. Zvi Y. Gitelman and Yaakov Ro’i, eds., Revolution, Repression, and Revival: the
Soviet Jewish Experience (New York: Rowman & Littlefield, 2007), 119.

38. Frederic Bozo, Marie-Pierre Rey, N. Piers Ludlow, and Bernd Rother, eds., Visions
of the End of the Cold War in Europe, 1945–1990 (New York: Berghahn Books, 2012),

39. Daniel Johnson, White King and Red Queen: How the Cold War Was Fought on the
Chessboard (Boston: Houghton Mifflin, 2008), esp. chap. 6.

40. Boris Stillman, Linguistic Geometry (New York: Springer, 2000), xi.

41. Bruce Abramson, Digital Phoenix: Why the Information Economy Collapsed and How
It Will Rise Again (Cambridge: MIT Press), 89–90.

42. Johnson, White King and Red Queen, chap. 6.

43. Nathan Engsmenger, “Is Chess the Drosophila of Artificial Intelligence?,” Social
Studies of Science 42 (1) (2011): 5–30. See also John McCarthy, “Chess as the Dro-
sophila of AI,” accessed April 15, 2015, http://jmc.stanford.edu/articles/drosophila/

44. E. M. Landis and I. M. Yaglom, “About Aleksandr Semenovich Kronrod,” Uspekhi

Matematicheskikh Nauk 56 (5) (2001): 191–201, accessed April 15, 2015, http://www.

45. Wiener, God and Golem, Inc., 15–25.

46. Jad Abumrad and Robert Krulwich (hosts), “The Rules Can Set You Free,” Radio-
Lab, National Public Radio, April 9, 2013.

47. Walter Ong, Orality and Literacy (New York: Routledge, [1972] 2012), 82.

48. Philip von Hilger, War Games: A History of War on Paper (Cambridge: MIT Press,

49. Viktor Glushkov and V. Ya. Valakh, Chto takoe OGAS? (Moscow: Hauka, 1981),

9800.indb 254 6/2/16 3:05 PM

Notes to Conclusion  255

50. Malinovsky, Vechno Khranit, 57–58.

51. Letter from A. I. Kitov written on November 11, 1985, Politechnicheskii museum
Russian Funderation, fond “Kitov Anatolii Ivanovich,” f. 228, box BP 3450/1–2,
reproduced in the appendix to Kuteinikov, “Project Obshchegosudarsvennoi.”

52. Ibid.

53. Fedorenko, Vspominaya Proshloe, Vzgladivaya v Budushchee, 177.

54. Gerovitch, From Newspeak to Cyberspeak, 277.

55. Gerovitch, “Soviet InterNyet,” 346; V. Golovachev, “A Hercules Is Born,” Soviet

Cybernetics: Recent News Items 5 (1967): 72.

56. Gerovitch, From Newspeak to Cyberspeak, 139.

57. Graham, Lonely Ideas: Can Russia Compete?

58. Gregory, Restructuring the Soviet Economic Bureaucracy, esp. introduction.

59. Kuteinikov, “Proyekt obshchegosudarstvennoi,” 142.

60. Steven G. Medema, The Hesitant Hand: Taming Self-Interest in the History of Eco-
nomic Ideas (Princeton: Princeton University Press, 2010), 6–10. See also Pierre Force,
Self-Interest before Adam Smith: A Genealogy of Economic Science (New York: Cambridge
University Press, 2003), which sees self-interest as a first principle behind what
Hume calls the “selfish hypothesis” from the Epicureans through Jean Baptiste-Say.

61. Karl Eugen Wädekin, The Private Sector in Soviet Agriculture (Berkeley: University